aid stringlengths 9 15 | mid stringlengths 7 10 | abstract stringlengths 78 2.56k | related_work stringlengths 92 1.77k | ref_abstract dict |
|---|---|---|---|---|
1812.10037 | 2906637185 | Semantic parsing is the task of converting natural language utterances into machine interpretable meaning representations which can be executed against a real-world environment such as a database. Scaling semantic parsing to arbitrary domains faces two interrelated challenges: obtaining broad coverage training data eff... | The next breakthrough came with the work of zettlemoyer:learning:2005 , who introduced CCG in semantic parsing. Their probabilistic CCG grammars can deal with long range dependencies and construct non-projective meaning representations. A great deal of work follows zettlemoyer:learning:2005 but focuses on more fine-gra... | {
"cite_N": [
"@cite_30",
"@cite_35",
"@cite_42",
"@cite_44",
"@cite_43",
"@cite_15",
"@cite_75",
"@cite_20",
"@cite_8",
"@cite_52",
"@cite_49",
"@cite_80",
"@cite_70",
"@cite_55",
"@cite_25",
"@cite_33",
"@cite_53",
"@cite_63",
"@cite_13"
... |
1812.10071 | 2906542368 | Many semantic video analysis tasks can benefit from multiple, heterogenous signals. For example, in addition to the original RGB input sequences, sequences of optical flow are usually used to boost the performance of human action recognition in videos. To learn from these heterogenous input sources, existing methods re... | Countless learning tasks require dealing with sequential data. Image captioning @cite_0 , speech synthesis, and music generation all require that a model produce outputs that are sequences. In other domains, such as time series prediction, video analysis @cite_35 , and musical information retrieval, a model must learn ... | {
"cite_N": [
"@cite_0",
"@cite_35"
],
"mid": [
"1811254738",
"2182762369"
],
"abstract": [
"In this paper, we present a multimodal Recurrent Neural Network (m-RNN) model for generating novel image captions. It directly models the probability distribution of generating a word given previou... |
1812.10071 | 2906542368 | Many semantic video analysis tasks can benefit from multiple, heterogenous signals. For example, in addition to the original RGB input sequences, sequences of optical flow are usually used to boost the performance of human action recognition in videos. To learn from these heterogenous input sources, existing methods re... | Some LSTM-based two-stream networks have been proposed as well. @cite_16 @cite_18 proposed to train video recognition models using LSTMs that capture temporal state dependencies and explicitly model short snippets of ConvNet activations. @cite_8 demonstrated that two-stream LSTMs outperform improved dense trajectories ... | {
"cite_N": [
"@cite_18",
"@cite_26",
"@cite_8",
"@cite_29",
"@cite_1",
"@cite_3",
"@cite_16"
],
"mid": [
"",
"2963447094",
"1923404803",
"2156303437",
"2963246338",
"2105101328",
"2951183276"
],
"abstract": [
"",
"Human actions captured in video... |
1812.10071 | 2906542368 | Many semantic video analysis tasks can benefit from multiple, heterogenous signals. For example, in addition to the original RGB input sequences, sequences of optical flow are usually used to boost the performance of human action recognition in videos. To learn from these heterogenous input sources, existing methods re... | Not just action recognition, other computer vision tasks also consider using multiple branches to improve the accuracy. Based on the multi-stage work of @cite_7 , @cite_17 presents a real-time pose estimation method. They add a bottom-up representation of association scores via part affinity fields (PAFs). By adding th... | {
"cite_N": [
"@cite_7",
"@cite_17"
],
"mid": [
"2964304707",
"2951856387"
],
"abstract": [
"Pose Machines provide a sequential prediction framework for learning rich implicit spatial models. In this work we show a systematic design for how convolutional networks can be incorporated into t... |
1812.09832 | 2905751765 | Despite the significant success in image-to-image translation and latent representation based facial attribute editing and expression synthesis, the existing approaches still have limitations in the sharpness of details, distinct image translation and identity preservation. To address these issues, we propose a Texture... | @cite_2 is a novel generative model which decomposes the input image into texture and deformation. DAE follows the deformable template paradigm and models image generation through texture synthesis and spatial deformation. DAE can obtain the prototypical object by removing the deformation. Discarding variability due to... | {
"cite_N": [
"@cite_2"
],
"mid": [
"2807725536"
],
"abstract": [
"In this work we introduce Deforming Autoencoders, a generative model for images that disentangles shape from appearance in an unsupervised manner. As in the deformable template paradigm, shape is represented as a deformation betwee... |
1812.09832 | 2905751765 | Despite the significant success in image-to-image translation and latent representation based facial attribute editing and expression synthesis, the existing approaches still have limitations in the sharpness of details, distinct image translation and identity preservation. To address these issues, we propose a Texture... | @cite_11 is a promising generative model and can be used to solve various computer vision tasks such as image generation @cite_18 @cite_7 @cite_22 , image translation @cite_17 @cite_5 @cite_9 , and face image editing @cite_16 @cite_25 @cite_21 . The GAN model is mainly designed to learn a generator G to generate fake s... | {
"cite_N": [
"@cite_18",
"@cite_11",
"@cite_22",
"@cite_7",
"@cite_9",
"@cite_21",
"@cite_5",
"@cite_16",
"@cite_25",
"@cite_20",
"@cite_17"
],
"mid": [
"2952010110",
"",
"",
"2963567641",
"2608015370",
"2797823148",
"2962793481",
"29641... |
1812.09832 | 2905751765 | Despite the significant success in image-to-image translation and latent representation based facial attribute editing and expression synthesis, the existing approaches still have limitations in the sharpness of details, distinct image translation and identity preservation. To address these issues, we propose a Texture... | @cite_17 is a typical image-to-image translation based method. The approach can learn the mapping between input and output domains and has achieved impressive results in several image translation tasks @cite_5 @cite_9 @cite_24 . Pix2Pix combines adversarial loss with L1 loss to transfer images in a paired way. For unpa... | {
"cite_N": [
"@cite_9",
"@cite_24",
"@cite_0",
"@cite_5",
"@cite_13",
"@cite_17"
],
"mid": [
"2608015370",
"",
"2797650215",
"2962793481",
"2552611751",
""
],
"abstract": [
"Conditional Generative Adversarial Networks (GANs) for cross-domain image-to-image ... |
1812.09832 | 2905751765 | Despite the significant success in image-to-image translation and latent representation based facial attribute editing and expression synthesis, the existing approaches still have limitations in the sharpness of details, distinct image translation and identity preservation. To address these issues, we propose a Texture... | @cite_20 is a multiple facial attribute editing model that contains three components at training: the attribute classification constraint, the reconstruction learning and the adversarial learning. The content that latent representation deliveries is uncertain and limited. Hence, imposing the attribute label to the late... | {
"cite_N": [
"@cite_20"
],
"mid": [
"2770587987"
],
"abstract": [
"Facial attribute editing aims to modify either single or multiple attributes on a face image. Since it is practically infeasible to collect images with arbitrarily specified attributes for each person, the generative adversarial n... |
1812.09551 | 2952522726 | Taxonomy construction is not only a fundamental task for semantic analysis of text corpora, but also an important step for applications such as information filtering, recommendation, and Web search. Existing pattern-based methods extract hypernym-hyponym term pairs and then organize these pairs into a taxonomy. However... | There have also been (semi-)supervised learning methods for taxonomy construction @cite_25 @cite_7 . Basically these methods extract lexical features and learn a classifier that categorizes term pairs into relations or non-relations, based on curated training data of hypernym-hyponym pairs @cite_28 @cite_2 @cite_33 @ci... | {
"cite_N": [
"@cite_33",
"@cite_7",
"@cite_28",
"@cite_21",
"@cite_32",
"@cite_17",
"@cite_6",
"@cite_24",
"@cite_2",
"@cite_34",
"@cite_25",
"@cite_20",
"@cite_11"
],
"mid": [
"1603863883",
"38703128",
"2155734303",
"2138605095",
"151650166... |
1812.09670 | 2964162217 | Detecting vehicles with strong robustness and high efficiency has become one of the key capabilities of fully autonomous driving cars. This topic has already been widely studied by GPU-accelerated deep learning approaches using image sensors and 3D LiDAR, however, few studies seek to address it with a horizontally moun... | Also some other approaches were developed using volumetric data with 3D LiDARs, among which some choose sequential projections of point clouds @cite_6 , @cite_9 , others choose to train up neural networks that can cope with unordered point cloud data with abstract feature learning, like in VoxelNet and PointNet. Howeve... | {
"cite_N": [
"@cite_9",
"@cite_6"
],
"mid": [
"2337890890",
"2211722331"
],
"abstract": [
"In this work we present a novel end-to-end framework for tracking and classifying a robot's surroundings in complex, dynamic and only partially observable real-world environments. The approach deplo... |
1812.09537 | 2906641606 | Motivation: Traditional computational cluster schedulers are based on user inputs and run time needs request for memory and CPU, not IO. Heavily IO bound task run times, like ones seen in many big data and bioinformatics problems, are dependent on the IO subsystems scheduling and are problematic for cluster resource sc... | OVIS @cite_11 has attempted to address scheduling of resources based on predictive failure analysis but not on the user requested resource allocation. Additionally, OVIS's scope was limited to resource allocation improvements to address task scheduling around node failures not optimizing cluster resources. While a give... | {
"cite_N": [
"@cite_11"
],
"mid": [
"2162485312"
],
"abstract": [
"Traditional cluster monitoring approaches consider nodes in singleton, using manufacturer-specified extreme limits as thresholds for failure \"prediction\". We have developed a tool, OVIS, for monitoring and analysis of large comp... |
1812.09537 | 2906641606 | Motivation: Traditional computational cluster schedulers are based on user inputs and run time needs request for memory and CPU, not IO. Heavily IO bound task run times, like ones seen in many big data and bioinformatics problems, are dependent on the IO subsystems scheduling and are problematic for cluster resource sc... | Larger scale system monitoring of HPC and the Open Science Grid @cite_24 have used such systems as OVIS @cite_11 or TACC Stats @cite_1 . OVIS uses a Bayesian inference scheme to dynamically infer models for the normal behavior of a system and to determine bounds on the probability of values evinced in the system. OVIS ... | {
"cite_N": [
"@cite_24",
"@cite_1",
"@cite_11"
],
"mid": [
"2124088880",
"2011516616",
"2162485312"
],
"abstract": [
"The Open Science Grid (OSG) provides a distributed facility where the Consortium members provide guaranteed and opportunistic access to shared computing and storag... |
1812.09380 | 2905888739 | Recommendation systems are widely used by different user service providers specially those who have interactions with the large community of users. This paper introduces a recommender system based on community detection. The recommendation is provided using the local and global similarities between users. The local inf... | Ghavipour and Meybodi @cite_9 has introduced a fuzzy method in which, the level of users trust to each other is defined as fuzzy. For this purpose, they propose a method to adjust membership functions of fuzzy trust and distrust in recommender systems by using learning automata. | {
"cite_N": [
"@cite_9"
],
"mid": [
"2531287176"
],
"abstract": [
"We propose a learning automata-based method for optimizing membership functions.The proposed method adjusts the number and the position of membership functions.The proposed method can be used without any change in any fuzzy recomme... |
1812.09380 | 2905888739 | Recommendation systems are widely used by different user service providers specially those who have interactions with the large community of users. This paper introduces a recommender system based on community detection. The recommendation is provided using the local and global similarities between users. The local inf... | @cite_6 presented a clustering algorithm to solve gray-sheep users problem in recommender systems. They demonstrated that collaborative filtering algorithms fail to make accurate recommendations for gray-sheep users, so they proposed k-means clustering algorithm to identify these users and make reliable recommendations... | {
"cite_N": [
"@cite_6"
],
"mid": [
"1974196922"
],
"abstract": [
"We provide detailed analysis of gray-sheep users problem in recommender systems.We show how conventional collaborative filtering fail for gray-sheep users problem.We use K-means clustering to separate these users from rest of the u... |
1812.09380 | 2905888739 | Recommendation systems are widely used by different user service providers specially those who have interactions with the large community of users. This paper introduces a recommender system based on community detection. The recommendation is provided using the local and global similarities between users. The local inf... | Fulan @cite_4 proposed a new Community-based User domain Collaborative Recommendation Algorithm (CUCRA). This algorithm is performed in two section: firstly, it builds the offline user domain model; secondly, it recommends items to target users in the model by applying collaborative filtering. The former section consis... | {
"cite_N": [
"@cite_4"
],
"mid": [
"1558143660"
],
"abstract": [
"Collaborative Filtering (CF) is a commonly used technique in recommendation systems. It can promote items of interest to a target user from a large selection of available items. It is divided into two broad classes: memory-based al... |
1812.09380 | 2905888739 | Recommendation systems are widely used by different user service providers specially those who have interactions with the large community of users. This paper introduces a recommender system based on community detection. The recommendation is provided using the local and global similarities between users. The local inf... | Guo @cite_17 proposed three various approaches from the point of view of preference modeling to alleviate data sparsity and cold start problem. Low accuracy and coverage also are the issues of recommender system which have insufficient ratings. So this work addresses these issues in his proposed method, too. Firstly, i... | {
"cite_N": [
"@cite_17"
],
"mid": [
"2070845054"
],
"abstract": [
"Our research aims to tackle the problems of data sparsity and cold start of traditional recommender systems. Insufficient ratings often result in poor quality of recommendations in terms of accuracy and coverage. To address these ... |
1812.09380 | 2905888739 | Recommendation systems are widely used by different user service providers specially those who have interactions with the large community of users. This paper introduces a recommender system based on community detection. The recommendation is provided using the local and global similarities between users. The local inf... | @cite_10 proposed a multi-view clustering method to address the low accuracy and coverage in clustering-based recommender systems. In this method, users are iteratively clustered from the views of both rating patterns and social trust relationships. | {
"cite_N": [
"@cite_10"
],
"mid": [
"2117346966"
],
"abstract": [
"Although demonstrated to be efficient and scalable to large-scale data sets, clustering-based recommender systems suffer from relatively low accuracy and coverage. To address these issues, we develop a multiview clustering method ... |
1812.09380 | 2905888739 | Recommendation systems are widely used by different user service providers specially those who have interactions with the large community of users. This paper introduces a recommender system based on community detection. The recommendation is provided using the local and global similarities between users. The local inf... | Alizade and Sheugh @cite_19 proposed a multi-view clustering based on Euclidean distance by combining similarity-based distances and trust-based distances. This method reduces low accuracy and coverage in cluster-based recommender systems. | {
"cite_N": [
"@cite_19"
],
"mid": [
"2403905660"
],
"abstract": [
"In recent years, collaborative filtering (CF) methods are important and widely accepted techniques are available for recommender systems. One of these techniques is user based that produces useful recommendations based on the simi... |
1812.09150 | 2905908122 | This paper is devoted to the stochastic approximation of entropically regularized Wasserstein distances between two probability measures, also known as Sinkhorn divergences. The semi-dual formulation of such regularized optimal transportation problems can be rewritten as a non-strongly concave optimisation problem. It ... | Obtaining limiting distributions for empirical Wasserstein distances when both @math and @math are absolutely continuous measures has been the subject of various works in asymptotic statistics @cite_31 @cite_25 @cite_30 @cite_22 @cite_5 . For probability measures supported on finite spaces, limiting distributions for e... | {
"cite_N": [
"@cite_30",
"@cite_38",
"@cite_22",
"@cite_8",
"@cite_15",
"@cite_5",
"@cite_31",
"@cite_25"
],
"mid": [
"2611625744",
"2530707152",
"2026953971",
"2770187377",
"2897678613",
"2244432928",
"2082531187",
"2084887650"
],
"abstract": [... |
1812.09134 | 2808513732 | In this work, we propose a novel mobile rescue robot equipped with an immersive stereoscopic teleperception and a teleoperation control. This robot is designed with the capability to perform safely a casualty-extraction procedure. We have built a proof-of-concept mobile rescue robot called ResQbot for the experimental ... | Wide-ranging robotics research studies have been undertaken in the area of search, exploration, and monitoring, specifically with applications in SAR scenarios @cite_9 @cite_2 @cite_6 . Despite the use of the term 'rescue' in SAR, little attention has been given to the development of a rescue robot that is capable of p... | {
"cite_N": [
"@cite_9",
"@cite_6",
"@cite_2"
],
"mid": [
"1996985406",
"2107664584",
"1969161726"
],
"abstract": [
"In this paper, we propose a stochastic differential equation-based exploration algorithm to enable exploration in three-dimensional indoor environments with a payloa... |
1812.09355 | 2905381038 | Despite the remarkable evolution of deep neural networks in natural language processing (NLP), their interpretability remains a challenge. Previous work largely focused on what these models learn at the representation level. We break this analysis down further and study individual dimensions (neurons) in the vector rep... | Much of the previous work has looked into neural models from the perspective of what they learn about various language properties. This includes analyzing word and sentence embeddings @cite_6 @cite_23 @cite_5 , recurrent neural network (RNN) states @cite_3 @cite_27 , and NMT representations @cite_26 @cite_22 @cite_8 . ... | {
"cite_N": [
"@cite_26",
"@cite_22",
"@cite_8",
"@cite_1",
"@cite_3",
"@cite_6",
"@cite_27",
"@cite_23",
"@cite_5",
"@cite_12"
],
"mid": [
"2605717780",
"2773956126",
"2773621464",
"",
"2563574619",
"2515741950",
"2601836666",
"",
"27991... |
1812.09276 | 2951474242 | With the fast growth in the visual surveillance and security sectors, thermal infrared images have become increasingly necessary ina large variety of industrial applications. This is true even though IR sensors are still more expensive than their RGB counterpart having the same resolution. In this paper, we propose a d... | . Since super-resolution output is similar to the low-resolution input,with the high-frequency information missing, the learning can be made to produce only the residual information. VDSR @cite_1 and DRSN @cite_17 trained a model that learns residual information between LR and HR images. They used a skip connection, th... | {
"cite_N": [
"@cite_21",
"@cite_1",
"@cite_17"
],
"mid": [
"",
"2951997238",
"2949079773"
],
"abstract": [
"",
"We present a highly accurate single-image super-resolution (SR) method. Our method uses a very deep convolutional network inspired by VGG-net used for ImageNet class... |
1812.09276 | 2951474242 | With the fast growth in the visual surveillance and security sectors, thermal infrared images have become increasingly necessary ina large variety of industrial applications. This is true even though IR sensors are still more expensive than their RGB counterpart having the same resolution. In this paper, we propose a d... | The optimization procedure seeks to minimize the distance between the original HR image and the generated SR image. The most used optimization function in the SR problem is the content loss, which is done using the MSE as in @cite_1 or Charbonnier as in @cite_21 . SRGAN @cite_26 instead uses the adversarial loss and @c... | {
"cite_N": [
"@cite_21",
"@cite_1",
"@cite_12",
"@cite_26"
],
"mid": [
"",
"2951997238",
"2950689937",
"2523714292"
],
"abstract": [
"",
"We present a highly accurate single-image super-resolution (SR) method. Our method uses a very deep convolutional network inspired ... |
1812.09336 | 2905699315 | In this project, we worked on speech recognition, specifically predicting individual words based on both the video frames and audio. Empowered by convolutional neural networks, the recent speech recognition and lip reading models are comparable to human level performance. We re-implemented and made derivations of the s... | Earlier solutions to speech recognition mostly used either classical signal processing techniques or deep learning on only the video data or audio data to do the actual recognition. In the video space, LipNet @cite_33 is one example where a CNN is used with bi-directional GRU's to predict the word being said in the cur... | {
"cite_N": [
"@cite_32",
"@cite_4",
"@cite_33"
],
"mid": [
"2596627958",
"2594690981",
"2578229578"
],
"abstract": [
"We propose an end-to-end deep learning architecture for word-level visual speech recognition. The system is a combination of spatiotemporal convolutional, residual... |
1812.09336 | 2905699315 | In this project, we worked on speech recognition, specifically predicting individual words based on both the video frames and audio. Empowered by convolutional neural networks, the recent speech recognition and lip reading models are comparable to human level performance. We re-implemented and made derivations of the s... | Deep architectures that use both audio and video data also tend to use LSTM or GRU units for their predictions. This is seen in the encoder-decoder architecture employed in @cite_20 which uses unidirectional LSTMs to encode both the image and audio data and generates attention vectors to predict the word being said. Wh... | {
"cite_N": [
"@cite_15",
"@cite_13",
"@cite_32",
"@cite_20"
],
"mid": [
"2787944098",
"2474638510",
"2596627958",
"2952746495"
],
"abstract": [
"Several end-to-end deep learning approaches have been recently presented which extract either audio or visual features from the ... |
1812.08927 | 2906528293 | Two-sample testing is a fundamental problem in statistics. Despite its long history, there has been renewed interest in this problem with the advent of high-dimensional and complex data. Specifically, in the machine learning literature, there have been recent methodological developments such as classification accuracy ... | In recent years, several attempts have been made to connect binary classification with two-sample testing. The main idea of this approach is to check whether the accuracy of a binary classifier is better than chance level and reject the null if the difference is significant. Such an approach, referred to as an accuracy... | {
"cite_N": [
"@cite_7",
"@cite_8"
],
"mid": [
"362526619",
"2187584467"
],
"abstract": [
"It is shown how classification learning machines can be used to do multivariate goodness-of-fit and two-sample testing.",
"Comparing two sets of multivariate samples is a central problem in data ... |
1812.08781 | 2903787679 | We study object recognition under the constraint that each object class is only represented by very few observations. Semi-supervised learning, transfer learning, and few-shot recognition all concern with achieving fast generalization with few labeled data. In this paper, we propose a generic framework that utilizes un... | To solve a computer vision problem, it has become a common practice to build a large-scale dataset @cite_3 @cite_6 and train deep neural networks @cite_0 @cite_8 on it. This philosophy has achieved unprecedented success on many important computer vision problems @cite_3 @cite_23 @cite_26 . However, constructing a large... | {
"cite_N": [
"@cite_26",
"@cite_8",
"@cite_6",
"@cite_3",
"@cite_0",
"@cite_23"
],
"mid": [
"2117539524",
"1686810756",
"2108598243",
"2031489346",
"2163605009",
"1861492603"
],
"abstract": [
"The ImageNet Large Scale Visual Recognition Challenge is a bench... |
1812.08781 | 2903787679 | We study object recognition under the constraint that each object class is only represented by very few observations. Semi-supervised learning, transfer learning, and few-shot recognition all concern with achieving fast generalization with few labeled data. In this paper, we propose a generic framework that utilizes un... | Semi-supervised learning @cite_42 is a problem that lies in between supervised learning and unsupervised learning. It aims to make more accurate predictions by leveraging a large amount of unlabeled data than by relying on the labeled data alone. In the era of deep learning, one line of work leverages unlabeled data th... | {
"cite_N": [
"@cite_35",
"@cite_4",
"@cite_9",
"@cite_42",
"@cite_43",
"@cite_27",
"@cite_5",
"@cite_31",
"@cite_16",
"@cite_25"
],
"mid": [
"",
"2074668987",
"2949416428",
"2407712691",
"2606711863",
"2951970475",
"830076066",
"2122457239",... |
1812.08781 | 2903787679 | We study object recognition under the constraint that each object class is only represented by very few observations. Semi-supervised learning, transfer learning, and few-shot recognition all concern with achieving fast generalization with few labeled data. In this paper, we propose a generic framework that utilizes un... | Given some training data in training categories, few-shot recognition @cite_28 requires the classifier to generalize to new categories from observing very few examples, often 1-shot or 5-shot. A body of work approaches this problem by offline metric learning @cite_32 @cite_39 @cite_18 , where a generic similarity metri... | {
"cite_N": [
"@cite_18",
"@cite_37",
"@cite_28",
"@cite_1",
"@cite_32",
"@cite_39",
"@cite_40",
"@cite_12",
"@cite_11"
],
"mid": [
"2949442616",
"2770468159",
"2144209400",
"2787035179",
"2963341924",
"2601450892",
"2951881474",
"2742093937",
... |
1812.08781 | 2903787679 | We study object recognition under the constraint that each object class is only represented by very few observations. Semi-supervised learning, transfer learning, and few-shot recognition all concern with achieving fast generalization with few labeled data. In this paper, we propose a generic framework that utilizes un... | Since the inception of the ImageNet challenge @cite_26 , transfer learning has emerged almost everywhere in visual recognition, such as in object detection @cite_21 and semantic segmentation @cite_29 , by simply transferring the network weights learned on ImageNet classification and finetuning on the target task. When ... | {
"cite_N": [
"@cite_26",
"@cite_41",
"@cite_29",
"@cite_21",
"@cite_10"
],
"mid": [
"2117539524",
"2767657961",
"1903029394",
"2102605133",
"2588646734"
],
"abstract": [
"The ImageNet Large Scale Visual Recognition Challenge is a benchmark in object category classi... |
1812.08848 | 2906075167 | The Saliency Model Implementation Library for Experimental Research (SMILER) is a new software package which provides an open, standardized, and extensible framework for maintaining and executing computational saliency models. This work drastically reduces the human effort required to apply saliency algorithms to new t... | It should be noted that the current collection of models supported by SMILER consists of models which focus on pixel-wise assignment of conspicuity values and which have been predominantly applied to the domain of human fixation prediction. There are, however, other branches of saliency model research, such as salient ... | {
"cite_N": [
"@cite_19",
"@cite_27",
"@cite_59",
"@cite_46",
"@cite_12"
],
"mid": [
"1970342316",
"2130502991",
"1782414784",
"",
"2164720308"
],
"abstract": [
"This paper presents a spatio-temporal saliency model that predicts eye movement during video free viewin... |
1812.09041 | 2906177770 | Emotional content is a crucial ingredient in user-generated videos. However, the sparsely expressed emotions in the user-generated video cause difficulties to emotions analysis in videos. In this paper, we propose a new neural approach---Bi-stream Emotion Attribution-Classification Network (BEAC-Net) to solve three rel... | Extensive research has been performed on the problem of recognizing emotions from visual information. Most work follow psychological theories that lay out a fixed number of emotion categories, such as Ekman's six pan-cultural basic emotions @cite_60 @cite_2 and Plutchik's wheel of emotion @cite_62 . These emotions are ... | {
"cite_N": [
"@cite_62",
"@cite_2",
"@cite_60"
],
"mid": [
"2321825897",
"2126181565",
"2046677541"
],
"abstract": [
"",
"How do emotions and moods color cognition? In this article, we examine how such reactions influence both judgments and cognitive performance. We argue that... |
1812.09041 | 2906177770 | Emotional content is a crucial ingredient in user-generated videos. However, the sparsely expressed emotions in the user-generated video cause difficulties to emotions analysis in videos. In this paper, we propose a new neural approach---Bi-stream Emotion Attribution-Classification Network (BEAC-Net) to solve three rel... | Dimensional theories of emotion @cite_11 @cite_31 @cite_65 characterize emotions as points in a multi-dimensional space. This direction is theoretically appealing as it allows richer emotion descriptions than the basic categories. Early work almost exclusively use the two dimensions of valence and arousal @cite_11 , wh... | {
"cite_N": [
"@cite_38",
"@cite_22",
"@cite_7",
"@cite_65",
"@cite_23",
"@cite_31",
"@cite_11"
],
"mid": [
"2044807399",
"2766925079",
"2124801089",
"1999937463",
"2765291577",
"2156848952",
"2149628368"
],
"abstract": [
"Research in affective compu... |
1812.09041 | 2906177770 | Emotional content is a crucial ingredient in user-generated videos. However, the sparsely expressed emotions in the user-generated video cause difficulties to emotions analysis in videos. In this paper, we propose a new neural approach---Bi-stream Emotion Attribution-Classification Network (BEAC-Net) to solve three rel... | Various researchers explored features for visual emotion recognition, such as features enlightened by psychology and art theory @cite_46 and shape features @cite_24 . A classifier such as a support vector machine (SVM) or K-nearest neighbors (KNN) is trained to distinguish video's emotions. @cite_14 adapted a variant o... | {
"cite_N": [
"@cite_14",
"@cite_39",
"@cite_24",
"@cite_46",
"@cite_25"
],
"mid": [
"2156709807",
"2017411072",
"2085940040",
"2003856922",
"2146104196"
],
"abstract": [
"Affective understanding of film plays an important role in sophisticated movie analysis, ranki... |
1812.09041 | 2906177770 | Emotional content is a crucial ingredient in user-generated videos. However, the sparsely expressed emotions in the user-generated video cause difficulties to emotions analysis in videos. In this paper, we propose a new neural approach---Bi-stream Emotion Attribution-Classification Network (BEAC-Net) to solve three rel... | Since facial expressions are important expressions of emotion, many researchers focused on recognizing emotions from facial expressions. @cite_20 paid close attention to viewers' facial signals for detection. @cite_45 extracted viewers' facial activities frame by frame and drew an emotional curve to classify each video... | {
"cite_N": [
"@cite_9",
"@cite_37",
"@cite_45",
"@cite_20"
],
"mid": [
"2134860945",
"2339620988",
"2120856140",
"2161809425"
],
"abstract": [
"Facial expression is temporally dynamic event which can be decomposed into a set of muscle motions occurring in different facial ... |
1812.09041 | 2906177770 | Emotional content is a crucial ingredient in user-generated videos. However, the sparsely expressed emotions in the user-generated video cause difficulties to emotions analysis in videos. In this paper, we propose a new neural approach---Bi-stream Emotion Attribution-Classification Network (BEAC-Net) to solve three rel... | Deep neural networks have also been used for visual sentiment analysis @cite_56 @cite_58 . A massive scale of visual sentiment dataset was proposed in Sentibank @cite_58 and DeepSentiBank @cite_55 . Sentibank is composed of 1,533 adjective-noun pairs, such as happy dog'' and beautiful sky''. Subsequently, the authors u... | {
"cite_N": [
"@cite_55",
"@cite_58",
"@cite_56"
],
"mid": [
"1784731433",
"2075456404",
"2963992782"
],
"abstract": [
"This paper introduces a visual sentiment concept classification method based on deep convolutional neural networks (CNNs). The visual sentiment concepts are adjec... |
1812.09041 | 2906177770 | Emotional content is a crucial ingredient in user-generated videos. However, the sparsely expressed emotions in the user-generated video cause difficulties to emotions analysis in videos. In this paper, we propose a new neural approach---Bi-stream Emotion Attribution-Classification Network (BEAC-Net) to solve three rel... | The emotional content in videos can be recognized from visual features, audio features and their combination. A number of work attempted to recognize emotionsand affects from speech @cite_29 @cite_47 @cite_13 . @cite_26 jointly uses speech and facial expressions. @cite_12 extracts mid-level audio-visual features. @cite... | {
"cite_N": [
"@cite_30",
"@cite_61",
"@cite_26",
"@cite_8",
"@cite_48",
"@cite_29",
"@cite_43",
"@cite_47",
"@cite_13",
"@cite_12"
],
"mid": [
"1930223417",
"2053233027",
"2168053878",
"2081835714",
"2210641851",
"2110052520",
"2173163709",
... |
1812.09041 | 2906177770 | Emotional content is a crucial ingredient in user-generated videos. However, the sparsely expressed emotions in the user-generated video cause difficulties to emotions analysis in videos. In this paper, we propose a new neural approach---Bi-stream Emotion Attribution-Classification Network (BEAC-Net) to solve three rel... | Most existing work focus on emotion understanding from video focus on classification. As emotional content are sparsely expressed in user-generated video, the task of identifying emotional segments in the video @cite_1 @cite_35 @cite_54 may provide assistance to the classification task. Noting the synergy between the t... | {
"cite_N": [
"@cite_35",
"@cite_54",
"@cite_1"
],
"mid": [
"2414501075",
"2177696193",
"2098287351"
],
"abstract": [
"Despite growing research interest, emotion understanding for user-generated videos remains a challenging problem. Major obstacles include the diversity and complex... |
1812.09041 | 2906177770 | Emotional content is a crucial ingredient in user-generated videos. However, the sparsely expressed emotions in the user-generated video cause difficulties to emotions analysis in videos. In this paper, we propose a new neural approach---Bi-stream Emotion Attribution-Classification Network (BEAC-Net) to solve three rel... | Video summarization has been studied for more than two decades @cite_3 and a detailed review is beyond the scope of this paper. In broad strokes, we can categorize summarization approaches into two major approaches: keyframes extraction and video skims. A large varienty of video features have been exploited, including ... | {
"cite_N": [
"@cite_18",
"@cite_4",
"@cite_28",
"@cite_3",
"@cite_0",
"@cite_10"
],
"mid": [
"2092825074",
"2006180404",
"2061833242",
"2094998392",
"2095536970",
"2014464472"
],
"abstract": [
"Rushes footages are considered as cheap gold mine with the pote... |
1812.09041 | 2906177770 | Emotional content is a crucial ingredient in user-generated videos. However, the sparsely expressed emotions in the user-generated video cause difficulties to emotions analysis in videos. In this paper, we propose a new neural approach---Bi-stream Emotion Attribution-Classification Network (BEAC-Net) to solve three rel... | Recently, we @cite_54 introduced the task of emotion-oriented summarization which points at finishing video summarization task according to video emotion content. Inspired by the task of semantic attribution in text analysis, the task of emotion attribution @cite_54 are defined as attributing the video's overall emotio... | {
"cite_N": [
"@cite_54"
],
"mid": [
"2177696193"
],
"abstract": [
"Emotion is a key element in user-generated video. However, it is difficult to understand emotions conveyed in such videos due to the complex and unstructured nature of user-generated content and the sparsity of video frames expres... |
1812.09041 | 2906177770 | Emotional content is a crucial ingredient in user-generated videos. However, the sparsely expressed emotions in the user-generated video cause difficulties to emotions analysis in videos. In this paper, we propose a new neural approach---Bi-stream Emotion Attribution-Classification Network (BEAC-Net) to solve three rel... | In the previous work @cite_41 , we only focused on the segment of high emotional value while neglected other frames which may contain content information. However, in this paper, the emotion segment and the entire content information will be combined with different emphasis. | {
"cite_N": [
"@cite_41"
],
"mid": [
"2617085328"
],
"abstract": [
"Emotional content is a key ingredient in user-generated videos. However, due to the emotion sparsely expressed in the user-generated video, it is very difficult to analayze emotions in videos. In this paper, we propose a new archi... |
1812.09041 | 2906177770 | Emotional content is a crucial ingredient in user-generated videos. However, the sparsely expressed emotions in the user-generated video cause difficulties to emotions analysis in videos. In this paper, we propose a new neural approach---Bi-stream Emotion Attribution-Classification Network (BEAC-Net) to solve three rel... | The proposed technique in this paper is inspired partially by the spatial transform network (ST-net) @cite_52 , which is firstly proposed for image (or feature map) classification. ST-net provides the capability for spatial transformation, which helps various tasks such as co-localization @cite_63 and spatial attention... | {
"cite_N": [
"@cite_5",
"@cite_52",
"@cite_63"
],
"mid": [
"2950178297",
"603908379",
""
],
"abstract": [
"Inspired by recent work in machine translation and object detection, we introduce an attention based model that automatically learns to describe the content of images. We des... |
1812.09041 | 2906177770 | Emotional content is a crucial ingredient in user-generated videos. However, the sparsely expressed emotions in the user-generated video cause difficulties to emotions analysis in videos. In this paper, we propose a new neural approach---Bi-stream Emotion Attribution-Classification Network (BEAC-Net) to solve three rel... | So far there are various variants and improvement of ST-net. @cite_33 adapted it for end-to-end facial learning framework and proposed a loss function for solving the problem that ST-net might lead to its output patch beyond the input boundaries. @cite_36 improved upon ST-net by theoretically connecting it to the inver... | {
"cite_N": [
"@cite_36",
"@cite_33"
],
"mid": [
"2562066862",
"2499554887"
],
"abstract": [
"In this paper, we establish a theoretical connection between the classical Lucas & Kanade (LK) algorithm and the emerging topic of Spatial Transformer Networks (STNs). STNs are of interest to the ... |
1812.09041 | 2906177770 | Emotional content is a crucial ingredient in user-generated videos. However, the sparsely expressed emotions in the user-generated video cause difficulties to emotions analysis in videos. In this paper, we propose a new neural approach---Bi-stream Emotion Attribution-Classification Network (BEAC-Net) to solve three rel... | BEAC-Net contains a two-stream architecture that extract features not only from the video segment identified by the attribution network, but also the entire video as its context. This is different from the two-stream architecture introduced by @cite_59 , which contains a convolutional stream to process pixels of the fr... | {
"cite_N": [
"@cite_50",
"@cite_59"
],
"mid": [
"2963524571",
"2156303437"
],
"abstract": [
"The paucity of videos in current action classification datasets (UCF-101 and HMDB-51) has made it difficult to identify good video architectures, as most methods obtain similar performance on exis... |
1812.08901 | 2906262207 | The acceptance of autonomous vehicles is dependent on the rigorous assessment of their safety. Furthermore, the commercial viability of AV programs depends on the ability to estimate the time and resources required to achieve desired safety levels. Naive approaches to estimating the reliability and safety levels of aut... | In this context, we should also consider the work of Huang al @cite_24 , who observe that testing driverless cars by reflecting real-world driving conditions is not a particularly efficient way to improve or measure reliability. Their approach is to consider particular driving tasks (in their example, freeway lane chan... | {
"cite_N": [
"@cite_24"
],
"mid": [
"2950211306"
],
"abstract": [
"The process to certify highly Automated Vehicles has not yet been defined by any country in the world. Currently, companies test Automated Vehicles on public roads, which is time-consuming and inefficient. We proposed the Accelera... |
1812.08901 | 2906262207 | The acceptance of autonomous vehicles is dependent on the rigorous assessment of their safety. Furthermore, the commercial viability of AV programs depends on the ability to estimate the time and resources required to achieve desired safety levels. Naive approaches to estimating the reliability and safety levels of aut... | @cite_4 examined accident data for the Waymo program through to 2017, and found a simple linear relationship between kilometres travelled and cumulative accidents. They therefore concluded, While accidents are an important metric, it does not follow that there have been no improvements in the function of AV systems. Th... | {
"cite_N": [
"@cite_4"
],
"mid": [
"2755893646"
],
"abstract": [
"Autonomous Vehicle technology is quickly expanding its market and has found in Silicon Valley, California, a strong foothold for preliminary testing on public roads. In an effort to promote safety and transparency to consumers, the... |
1812.08839 | 2906579741 | The ability to build a model on a source task and subsequently adapt such model on a new target task is a pervasive need in many astronomical applications. The problem is generally known as transfer learning in machine learning, where domain adaptation is a popular scenario. An example is to build a predictive model on... | Domain adaptation induces a model by exploiting experience gathered from previous tasks @cite_2 . It is considered a subfield of transfer learning @cite_3 , and has become increasingly popular in recent years due to the pervasive nature of task domains exhibiting differences in sample distribution @cite_35 @cite_7 . Th... | {
"cite_N": [
"@cite_35",
"@cite_7",
"@cite_3",
"@cite_2"
],
"mid": [
"2403788517",
"2951103356",
"2165698076",
"2104094955"
],
"abstract": [
"Domain adaptation aims at learning robust classifiers across domains using labeled data from a source domain. Representation learni... |
1812.08839 | 2906579741 | The ability to build a model on a source task and subsequently adapt such model on a new target task is a pervasive need in many astronomical applications. The problem is generally known as transfer learning in machine learning, where domain adaptation is a popular scenario. An example is to build a predictive model on... | Feature-based domain adaptation methods attempt to project source and target datasets into a latent feature space, where the covariate-shift assumption holds. A model is then built on the transformed space, and used as the classifier on the target. Examples are structural corresponding learning @cite_4 , subspace align... | {
"cite_N": [
"@cite_4",
"@cite_53",
"@cite_56",
"@cite_0",
"@cite_43"
],
"mid": [
"2158108973",
"2104068492",
"22861983",
"2130903752",
"2159570078"
],
"abstract": [
"Discriminative learning methods are widely used in natural language processing. These methods work... |
1812.08839 | 2906579741 | The ability to build a model on a source task and subsequently adapt such model on a new target task is a pervasive need in many astronomical applications. The problem is generally known as transfer learning in machine learning, where domain adaptation is a popular scenario. An example is to build a predictive model on... | From a theoretical view, previous work has tried to estimate the distance between source and target distributions @cite_5 @cite_2 @cite_8 ; and employ regularization terms to find models with good generalization performance on both source and target domains @cite_55 . | {
"cite_N": [
"@cite_5",
"@cite_55",
"@cite_8",
"@cite_2"
],
"mid": [
"2131953535",
"",
"2110091014",
"2104094955"
],
"abstract": [
"Discriminative learning methods for classification perform well when training and test data are drawn from the same distribution. In many sit... |
1812.08866 | 2906410737 | To support Machine Type Communications (MTC) in next generation mobile networks, NarrowBand-IoT (NB-IoT) has been released by the Third Generation Partnership Project (3GPP) as a promising solution to provide extended coverage and low energy consumption for low cost MTC devices. However, the existing Orthogonal Multipl... | Al-Imari @cite_23 proposed a NOMA scheme for uplink data transmission that allows multiple users to share the same sub-carrier without any coding spreading redundancy. Mostafa @cite_15 studied the connectivity maximization for the application of NOMA in NB-IoT, where only two users can share the same sub-carrier. Kiani... | {
"cite_N": [
"@cite_18",
"@cite_4",
"@cite_21",
"@cite_23",
"@cite_2",
"@cite_15",
"@cite_16",
"@cite_17"
],
"mid": [
"2963738831",
"2797309738",
"",
"1997958020",
"2897268940",
"2741981612",
"",
"2772709146"
],
"abstract": [
"Wireless power... |
1812.08972 | 2905752951 | There is recently a surge in approaches that learn low-dimensional embeddings of nodes in networks. As there are many large-scale real-world networks, it's inefficient for existing approaches to store amounts of parameters in memory and update them edge after edge. With the knowledge that nodes having similar neighborh... | Vertex classification is one of the most common semi-supervised tasks in network analysis, which aims to classify the vertices to at least one groups. The application of the task could be shown in many areas, such as protein classification @cite_25 , user profiling @cite_43 @cite_48 , and so on. | {
"cite_N": [
"@cite_43",
"@cite_25",
"@cite_48"
],
"mid": [
"2107961038",
"2962756421",
""
],
"abstract": [
"User attributes, such as occupation, education, and location, are important for many applications. In this paper, we study the problem of profiling user attributes in socia... |
1812.08972 | 2905752951 | There is recently a surge in approaches that learn low-dimensional embeddings of nodes in networks. As there are many large-scale real-world networks, it's inefficient for existing approaches to store amounts of parameters in memory and update them edge after edge. With the knowledge that nodes having similar neighborh... | Most current node embedding techniques are lookup algorithms, i.e., there is a matrix containing the embedding vectors for all nodes, so we just need to look up in the matrix for a specific embedding. Early works in NRL mainly are based on the factorization of the graph Laplacian matrix, such as Isomap @cite_23 , Lapla... | {
"cite_N": [
"@cite_62",
"@cite_21",
"@cite_3",
"@cite_6",
"@cite_23",
"@cite_63",
"@cite_58",
"@cite_25",
"@cite_20",
"@cite_17"
],
"mid": [
"2393319904",
"",
"2154851992",
"2173649752",
"",
"1888005072",
"2046253692",
"2962756421",
"25... |
1812.08972 | 2905752951 | There is recently a surge in approaches that learn low-dimensional embeddings of nodes in networks. As there are many large-scale real-world networks, it's inefficient for existing approaches to store amounts of parameters in memory and update them edge after edge. With the knowledge that nodes having similar neighborh... | Closely related to our model, HARP @cite_12 first coarsens the graph, and after that, the new graph consists of supernodes. Afterward, network embedding methods are applied to learn the representations of supernodes, and then with the learned representation as the initial value of the supernodes' constituent nodes, the... | {
"cite_N": [
"@cite_12",
"@cite_8"
],
"mid": [
"2700550412",
"2788760796"
],
"abstract": [
"We present HARP, a novel method for learning low dimensional embeddings of a graph's nodes which preserves higher-order structural features. Our proposed method achieves this by compressing the inp... |
1812.08972 | 2905752951 | There is recently a surge in approaches that learn low-dimensional embeddings of nodes in networks. As there are many large-scale real-world networks, it's inefficient for existing approaches to store amounts of parameters in memory and update them edge after edge. With the knowledge that nodes having similar neighborh... | Compression for Convolutional Neural Networks (CNN) has been extensively studied, mainly divided into three following branches. First, Low-rank matrix tensor factorization @cite_13 @cite_35 @cite_9 is derived on the assumption that using a low-rank approximation of the matrix to approximate each of the networks' weight... | {
"cite_N": [
"@cite_30",
"@cite_35",
"@cite_36",
"@cite_41",
"@cite_9",
"@cite_54",
"@cite_21",
"@cite_39",
"@cite_44",
"@cite_19",
"@cite_59",
"@cite_31",
"@cite_13",
"@cite_11"
],
"mid": [
"",
"2950967261",
"2119144962",
"2950894517",
... |
1812.08843 | 2949770501 | In important applications involving multi-task networks with multiple objectives, agents in the network need to decide between these multiple objectives and reach an agreement about which single objective to follow for the network. In this work we propose a distributed decision-making algorithm. The agents are assumed ... | Bio-inspired systems are designed to mimic the behavior of some animal groups such as bee swarms, birds flying in formation, and schools of fish @cite_11 @cite_0 @cite_17 @cite_10 @cite_9 @cite_16 . Diffusion strategies can be used to model some of these coordinated types of behavior, as well as solve inference and est... | {
"cite_N": [
"@cite_13",
"@cite_18",
"@cite_11",
"@cite_14",
"@cite_8",
"@cite_9",
"@cite_1",
"@cite_3",
"@cite_6",
"@cite_0",
"@cite_19",
"@cite_5",
"@cite_15",
"@cite_16",
"@cite_10",
"@cite_12",
"@cite_17"
],
"mid": [
"2046058958",
"1... |
1812.08843 | 2949770501 | In important applications involving multi-task networks with multiple objectives, agents in the network need to decide between these multiple objectives and reach an agreement about which single objective to follow for the network. In this work we propose a distributed decision-making algorithm. The agents are assumed ... | In the latter case, agents need to decide between the multiple objectives and reach agreement on following a single objective for the entire network. In the earlier works @cite_7 @cite_4 , a scenario was considered where agents were assumed to sense data arising from two models, and a diffusion strategy was developed t... | {
"cite_N": [
"@cite_4",
"@cite_7"
],
"mid": [
"2577808232",
"2112603423"
],
"abstract": [
"In this paper, we study distributed decision-making over mobile adaptive networks where nodes in the network collect data generated by two different models. The nodes need to decide which model to e... |
1812.08843 | 2949770501 | In important applications involving multi-task networks with multiple objectives, agents in the network need to decide between these multiple objectives and reach an agreement about which single objective to follow for the network. In this work we propose a distributed decision-making algorithm. The agents are assumed ... | We consider a distributed mean-square-error estimation problem over an @math -agent network. The connectivity of the agents is described by a graph (see Fig. ). Data sensed by any particular agent can arise from one of different models. The objective is to reach an agreement among all agents in the network on one commo... | {
"cite_N": [
"@cite_2"
],
"mid": [
"2545849206"
],
"abstract": [
"We consider the problem of decentralized clustering and estimation over multitask networks, where agents infer and track different models of interest. The agents do not know beforehand which model is generating their own data. They... |
1907.03112 | 2955263739 | Cross-lingual embeddings aim to represent words in multiple languages in a shared vector space by capturing semantic similarities across languages. They are a crucial component for scaling tasks to multiple languages by transferring knowledge from languages with rich resources to low-resource languages. A common approa... | A common way to select a bilingual dictionary is by using either automatic translations of frequent words or word alignments. For instance, @cite_1 select the target word to which the source word is most frequently aligned in parallel corpora. @cite_17 use the 5,000 most frequent words from the source language with the... | {
"cite_N": [
"@cite_1",
"@cite_4",
"@cite_17"
],
"mid": [
"342285082",
"2508069829",
"2126725946"
],
"abstract": [
"The distributional hypothesis of Harris (1954), according to which the meaning of words is evidenced by the contexts they occur in, has motivated several effective t... |
1907.03081 | 2954077899 | With the rise of the Internet of Things (IoT), fog computing has emerged to help traditional cloud computing in meeting scalability demands. Fog computing makes it possible to fulfill real-time requirements of applications by bringing more processing, storage, and control power geographically closer to edge devices. Ho... | CloudSim @cite_31 is perhaps the most popular cloud simulation platform available, which is used for modeling the cloud and application provisioning environments. It is a discrete, event-based simulator written in Java, meaning that it does not actually emulate network entities such as routers and switches. Instead, Cl... | {
"cite_N": [
"@cite_31"
],
"mid": [
"2045287414"
],
"abstract": [
"Cloud computing is a recent advancement wherein IT infrastructure and applications are provided as ‘services’ to end-users under a usage-based payment model. It can leverage virtualized services even on the fly based on requiremen... |
1907.03081 | 2954077899 | With the rise of the Internet of Things (IoT), fog computing has emerged to help traditional cloud computing in meeting scalability demands. Fog computing makes it possible to fulfill real-time requirements of applications by bringing more processing, storage, and control power geographically closer to edge devices. Ho... | There are also many extensions to CloudSim, such as CloudSimSDN @cite_21 , ContainerCloudSim @cite_33 , and iFogSim @cite_2 , which attempt to broaden CloudSim's model to include SDN, Docker container migration simulations, and fog computing, respectively. However, because CloudSim and these associated extensions are s... | {
"cite_N": [
"@cite_21",
"@cite_33",
"@cite_2"
],
"mid": [
"1512071441",
"2604186514",
"2414114959"
],
"abstract": [
"Software-Defined Networking not only addresses the shortcoming of traditional network technologies in dealing with frequent and immediate changes in cloud data cen... |
1907.03081 | 2954077899 | With the rise of the Internet of Things (IoT), fog computing has emerged to help traditional cloud computing in meeting scalability demands. Fog computing makes it possible to fulfill real-time requirements of applications by bringing more processing, storage, and control power geographically closer to edge devices. Ho... | Typically, network resource management is accomplished using a load balancer, which attempts to find a suitable path to one or more destinations while optimally spreading traffic throughout the network to avoid congestion. In many cases, Equal-Cost Multi-Path (ECMP) routing is used to manage network resources by distri... | {
"cite_N": [
"@cite_27",
"@cite_12"
],
"mid": [
"2744698795",
"2770706713"
],
"abstract": [
"Production datacenters operate under various uncertainties such as traffic dynamics, topology asymmetry, and failures. Therefore, datacenter load balancing schemes must be resilient to these uncer... |
1907.03081 | 2954077899 | With the rise of the Internet of Things (IoT), fog computing has emerged to help traditional cloud computing in meeting scalability demands. Fog computing makes it possible to fulfill real-time requirements of applications by bringing more processing, storage, and control power geographically closer to edge devices. Ho... | Resource allocation is key to the success of edge-fog systems, and many fog architectures involving automated resource allocation mechanisms have been proposed. Skarlat @cite_20 created a resource provisioning system for IoT services in fog networks using a fog-cloud middleware component. The middleware oversees the ac... | {
"cite_N": [
"@cite_20"
],
"mid": [
"2565437603"
],
"abstract": [
"The advent of the Internet of Things (IoT) leadsto the pervasion of business and private spaces with ubiquitous, networked computing devices. These devices do not simply actas sensors, but feature computational, storage, and netwo... |
1907.03081 | 2954077899 | With the rise of the Internet of Things (IoT), fog computing has emerged to help traditional cloud computing in meeting scalability demands. Fog computing makes it possible to fulfill real-time requirements of applications by bringing more processing, storage, and control power geographically closer to edge devices. Ho... | Yin @cite_10 built a novel task-scheduling algorithm and designed a resource reallocation algorithm for fog networks, specifically for real-time, smart manufacturing applications. However, unlike the previous work, a management software component is not used in their approach, and each fog node is burdened with the tas... | {
"cite_N": [
"@cite_10"
],
"mid": [
"2810048489"
],
"abstract": [
"Fog computing has been proposed as an extension of cloud computing to provide computation, storage, and network services in network edge. For smart manufacturing, fog computing can provide a wealth of computational and storage ser... |
1907.03081 | 2954077899 | With the rise of the Internet of Things (IoT), fog computing has emerged to help traditional cloud computing in meeting scalability demands. Fog computing makes it possible to fulfill real-time requirements of applications by bringing more processing, storage, and control power geographically closer to edge devices. Ho... | Finally, the work that is perhaps most similar to the FDK is ENORM: The Edge Node Resource Management framework by Wang @cite_29 . Upon startup of the system, an edge manager software installed on all edge nodes gathers and stores available system resources. Then, each edge node listens for resource requests from a clo... | {
"cite_N": [
"@cite_29"
],
"mid": [
"2755775376"
],
"abstract": [
"Current computing techniques using the cloud as a centralised server will become untenable as billions of devices get connected to the Internet. This raises the need for fog computing, which leverages computing at the edge of the ... |
1907.02908 | 2913403708 | Many deep reinforcement learning algorithms contain inductive biases that sculpt the agent's objective and its interface to the environment. These inductive biases can take many forms, including domain knowledge and pretuned hyper-parameters. In general, there is a trade-off between generality and performance when algo... | The present work was partially inspired by the work of @cite_1 in the context of Go. They demonstrated that specific domain specific heuristics (e.g. pretraining on human data, the use of handcrafted Go-specific features, and exploitation of certain symmetries in state space), while originally introduced to simplify le... | {
"cite_N": [
"@cite_1"
],
"mid": [
"2963403143"
],
"abstract": [
"The Arcade Learning Environment (ALE) is an evaluation platform that poses the challenge of building AI agents with general competency across dozens of Atari 2600 games. It supports a variety of different problem settings and it ha... |
1907.02908 | 2913403708 | Many deep reinforcement learning algorithms contain inductive biases that sculpt the agent's objective and its interface to the environment. These inductive biases can take many forms, including domain knowledge and pretuned hyper-parameters. In general, there is a trade-off between generality and performance when algo... | There are other two features of our algorithm that, despite not incorporating quite as much domain knowledge as the heuristics discussed in this paper, also constitute a potential impediment to its generality and scalability. 1) The use of parallel environments is not always feasible in practice, especially in real wor... | {
"cite_N": [
"@cite_5"
],
"mid": [
"2157864803"
],
"abstract": [
"Many practitioners of reinforcement learning problems have observed that oftentimes the performance of the agent reaches very close to the optimal performance even though the estimated (action-)value function is still far from the ... |
1907.02874 | 2953849475 | Sharing knowledge between tasks is vital for efficient learning in a multi-task setting. However, most research so far has focused on the easier case where knowledge transfer is not harmful, i.e., where knowledge from one task cannot negatively impact the performance on another task. In contrast, we present an approach... | Glatt al @cite_10 train a DQN on a source task and investigate how the learned weights, which are used as initialization for a target task, alter the performance. In a similar manner, @cite_30 @cite_13 @cite_15 show that some transfer is possible by simply training one network on multiple tasks. However, since these al... | {
"cite_N": [
"@cite_30",
"@cite_15",
"@cite_10",
"@cite_13"
],
"mid": [
"",
"2891076394",
"2585821313",
"2786036274"
],
"abstract": [
"",
"The reinforcement learning (RL) community has made great strides in designing algorithms capable of exceeding human performance on... |
1907.02874 | 2953849475 | Sharing knowledge between tasks is vital for efficient learning in a multi-task setting. However, most research so far has focused on the easier case where knowledge transfer is not harmful, i.e., where knowledge from one task cannot negatively impact the performance on another task. In contrast, we present an approach... | One interesting line of research @cite_8 @cite_7 @cite_16 @cite_11 @cite_9 capitalizes on transferring knowledge based on successor features, i.e., shared environment dynamics. In contrast, our method does not rely on shared environment dynamics nor action alignment across tasks. | {
"cite_N": [
"@cite_7",
"@cite_8",
"@cite_9",
"@cite_16",
"@cite_11"
],
"mid": [
"2963019567",
"2605369401",
"2914694967",
"2962717849",
"2951871955"
],
"abstract": [
"In this paper we consider the problem of robot navigation in simple maze-like environments where ... |
1907.02874 | 2953849475 | Sharing knowledge between tasks is vital for efficient learning in a multi-task setting. However, most research so far has focused on the easier case where knowledge transfer is not harmful, i.e., where knowledge from one task cannot negatively impact the performance on another task. In contrast, we present an approach... | Czarnecki al @cite_19 use multiple networks similar to our approach. However, their focus is on automated curriculum learning. Therefore they adjust the policy mixing weights through population based training @cite_23 while we learn attention weights conditioned on the task state. | {
"cite_N": [
"@cite_19",
"@cite_23"
],
"mid": [
"2803740478",
"2770298516"
],
"abstract": [
"We introduce MixM using our method to progress through an action-space curriculum we achieve both faster training and better final performance than one obtains using traditional methods. (2) We fu... |
1907.02874 | 2953849475 | Sharing knowledge between tasks is vital for efficient learning in a multi-task setting. However, most research so far has focused on the easier case where knowledge transfer is not harmful, i.e., where knowledge from one task cannot negatively impact the performance on another task. In contrast, we present an approach... | Rusu al @cite_17 introduce (PNN), an effective approach for learning in a sequential multi-task setting. In PNN, a new network and lateral connections for each additional task are added in order to enable knowledge transfer, which speeds up the training of subsequent tasks. The additional network parts let the architec... | {
"cite_N": [
"@cite_1",
"@cite_17"
],
"mid": [
"2963199420",
"2426267443"
],
"abstract": [
"Most deep reinforcement learning algorithms are data inefficient in complex and rich environments, limiting their applicability to many scenarios. One direction for improving data efficiency is mul... |
1907.03143 | 2954554213 | Knowledge graphs (KGs) typically contain temporal facts indicating relationships among entities at different times. Due to their incompleteness, several approaches have been proposed to infer new facts for a KG based on the existing ones-a problem known as KG completion. KG embedding approaches have proved effective fo... | Statistical relational AI (StaRAI) @cite_9 @cite_13 approaches are mainly based on soft (hanf-crafted or learned) rules @cite_1 @cite_26 @cite_57 @cite_62 where the probability of a world is typically proportional to the number of rules that are satisfied violated in that world and the confidence for each rule. A line ... | {
"cite_N": [
"@cite_30",
"@cite_37",
"@cite_26",
"@cite_62",
"@cite_33",
"@cite_14",
"@cite_7",
"@cite_9",
"@cite_1",
"@cite_32",
"@cite_57",
"@cite_19",
"@cite_45",
"@cite_2",
"@cite_13"
],
"mid": [
"1826836734",
"2606488140",
"1824971879",... |
1907.03143 | 2954554213 | Knowledge graphs (KGs) typically contain temporal facts indicating relationships among entities at different times. Due to their incompleteness, several approaches have been proposed to infer new facts for a KG based on the existing ones-a problem known as KG completion. KG embedding approaches have proved effective fo... | These approaches define weighted template walks on a KG and then answer queries by template matching @cite_5 @cite_4 . They have been shown to be quite similar to, and in some cases subsumed by, the models based on soft rules @cite_42 . | {
"cite_N": [
"@cite_5",
"@cite_42",
"@cite_4"
],
"mid": [
"2029249040",
"2789351899",
"1756422141"
],
"abstract": [
"Scientific literature with rich metadata can be represented as a labeled directed graph. This graph representation enables a number of scientific tasks such as ad h... |
1907.03143 | 2954554213 | Knowledge graphs (KGs) typically contain temporal facts indicating relationships among entities at different times. Due to their incompleteness, several approaches have been proposed to infer new facts for a KG based on the existing ones-a problem known as KG completion. KG embedding approaches have proved effective fo... | A large number of models have been developed for static KG embedding. A class of these models are the translational approaches corresponding to variations of TransE (see, e.g., @cite_49 @cite_15 @cite_3 ). Another class of approaches are based on a bilinear score function @math each imposing a different sparsity constr... | {
"cite_N": [
"@cite_28",
"@cite_48",
"@cite_58",
"@cite_3",
"@cite_6",
"@cite_40",
"@cite_49",
"@cite_23",
"@cite_63",
"@cite_15",
"@cite_34",
"@cite_17"
],
"mid": [
"2016753842",
"2145544171",
"2962850650",
"2463781041",
"2963432357",
"2964... |
1907.03143 | 2954554213 | Knowledge graphs (KGs) typically contain temporal facts indicating relationships among entities at different times. Due to their incompleteness, several approaches have been proposed to infer new facts for a KG based on the existing ones-a problem known as KG completion. KG embedding approaches have proved effective fo... | The idea behind our proposed embeddings is similar to diachronic word embeddings where a corpus is typically broken temporally into slices (e.g., 20-year chuncks of a 200-year corpus) and embeddings are learned for words in each chunk thus providing word embeddings that are a function of time (see, e.g., @cite_21 @cite... | {
"cite_N": [
"@cite_31",
"@cite_21",
"@cite_51",
"@cite_20"
],
"mid": [
"2416513196",
"2951300178",
"1570098300",
"2964231305"
],
"abstract": [
"Understanding how words change their meanings over time is key to models of language and cultural evolution, but historical data... |
1907.03141 | 2953797043 | Structured weight pruning is a representative model compression technique of DNNs to reduce the storage and computation requirements and accelerate inference. An automatic hyperparameter determination process is necessary due to the large number of flexible hyperparameters. This work proposes AutoSlim, an automatic str... | DNN weight pruning includes two major categories: the general, pruning @cite_31 @cite_41 @cite_7 @cite_11 @cite_12 @cite_2 where arbitrary weight can be pruned, and pruning @cite_13 @cite_31 @cite_32 @cite_16 @cite_39 that maintains certain regularity. Non-structured pruning can result in a higher pruning rate (weight ... | {
"cite_N": [
"@cite_7",
"@cite_41",
"@cite_32",
"@cite_39",
"@cite_2",
"@cite_31",
"@cite_16",
"@cite_13",
"@cite_12",
"@cite_11"
],
"mid": [
"2963674932",
"2507318699",
"2891561769",
"2884180697",
"",
"2701719801",
"2963363373",
"2513419314... |
1907.03141 | 2953797043 | Structured weight pruning is a representative model compression technique of DNNs to reduce the storage and computation requirements and accelerate inference. An automatic hyperparameter determination process is necessary due to the large number of flexible hyperparameters. This work proposes AutoSlim, an automatic str... | Figure illustrates three structured pruning schemes on the CONV layers of DNN: , , and (a.k.a. ), removing whole filter(s), channel(s), and the same location in each filter in each layer. CONV operations in DNNs are commonly transformed to matrix multiplications by converting weight tensors and feature map tensors to m... | {
"cite_N": [
"@cite_13"
],
"mid": [
"2513419314"
],
"abstract": [
"High demand for computation resources severely hinders deployment of large-scale Deep Neural Networks (DNN) in resource constrained devices. In this work, we propose a Structured Sparsity Learning (SSL) method to regularize the st... |
1907.03141 | 2953797043 | Structured weight pruning is a representative model compression technique of DNNs to reduce the storage and computation requirements and accelerate inference. An automatic hyperparameter determination process is necessary due to the large number of flexible hyperparameters. This work proposes AutoSlim, an automatic str... | It is also worth mentioning that filter pruning and channel pruning are correlated @cite_16 , as pruning a filter in layer @math (after batch norm) results in the removal of corresponding channel in layer @math . The relationship in ResNet @cite_9 and MobileNet @cite_22 will be more complicated due to bypass links. | {
"cite_N": [
"@cite_9",
"@cite_16",
"@cite_22"
],
"mid": [
"2194775991",
"2963363373",
"2963163009"
],
"abstract": [
"Deeper neural networks are more difficult to train. We present a residual learning framework to ease the training of networks that are substantially deeper than th... |
1907.03141 | 2953797043 | Structured weight pruning is a representative model compression technique of DNNs to reduce the storage and computation requirements and accelerate inference. An automatic hyperparameter determination process is necessary due to the large number of flexible hyperparameters. This work proposes AutoSlim, an automatic str... | Alternating Direction Method of Multipliers (ADMM) is a powerful mathematical optimization technique, by decomposing an original problem into two subproblems that can be solved separately and efficiently @cite_33 . Consider the general optimization problem @math . In ADMM, it is decomposed into two subproblems on @math... | {
"cite_N": [
"@cite_33"
],
"mid": [
"2164278908"
],
"abstract": [
"Many problems of recent interest in statistics and machine learning can be posed in the framework of convex optimization. Due to the explosion in size and complexity of modern datasets, it is increasingly important to be able to s... |
1907.03141 | 2953797043 | Structured weight pruning is a representative model compression technique of DNNs to reduce the storage and computation requirements and accelerate inference. An automatic hyperparameter determination process is necessary due to the large number of flexible hyperparameters. This work proposes AutoSlim, an automatic str... | As a key property, ADMM can effectively deal with a subset of combinatorial constraints and yield optimal (or at least high quality) solutions. The associated constraints in DNN weight pruning (both non-structured and structured) belong to this subset @cite_10 @cite_5 . In DNN weight pruning problem, @math is loss func... | {
"cite_N": [
"@cite_5",
"@cite_14",
"@cite_10",
"@cite_6"
],
"mid": [
"2765313807",
"1522301498",
"2295652899",
""
],
"abstract": [
"In this paper, we design and analyze a new zeroth-order online algorithm, namely, the zeroth-order online alternating direction method of mu... |
1907.03141 | 2953797043 | Structured weight pruning is a representative model compression technique of DNNs to reduce the storage and computation requirements and accelerate inference. An automatic hyperparameter determination process is necessary due to the large number of flexible hyperparameters. This work proposes AutoSlim, an automatic str... | Many recent work have investigated the concept of (AutoML), i.e., using machine learning for hyperparameter determination in DNNs. Neural architecture search (NAS) @cite_36 @cite_0 @cite_18 is an representative application of AutoML. NAS has been deployed in Google’s Cloud AutoML framework, which frees customers from t... | {
"cite_N": [
"@cite_35",
"@cite_18",
"@cite_36",
"@cite_28",
"@cite_1",
"@cite_0"
],
"mid": [
"",
"2771727678",
"2553303224",
"",
"2949941638",
"2951886768"
],
"abstract": [
"",
"We propose a new method for learning the structure of convolutional neural... |
1907.03030 | 2964184973 | This paper targets the problem of image set-based face verification and identification. Unlike traditional single media (an image or video) setting, we encounter a set of heterogeneous contents containing orderless images and videos. The importance of each image is usually considered either equal or based on their inde... | Reinforcement learning (RL) trains an agent to interact (by trail and error) with a dynamic environment with the objective to maximize its accumulated reward. Recently, deep RL with convolutional neural networks (CNN) achieved human-level performance in Atari Games @cite_12 . The CNN is an ideal approximate function to... | {
"cite_N": [
"@cite_28",
"@cite_53",
"@cite_42",
"@cite_65",
"@cite_44",
"@cite_12",
"@cite_20"
],
"mid": [
"2745868649",
"2950395671",
"2260756217",
"2121863487",
"2165150801",
"2145339207",
""
],
"abstract": [
"Deep reinforcement learning (DRL) is... |
1907.03030 | 2964184973 | This paper targets the problem of image set-based face verification and identification. Unlike traditional single media (an image or video) setting, we encounter a set of heterogeneous contents containing orderless images and videos. The importance of each image is usually considered either equal or based on their inde... | Besides, policy-based and actor-critic methods have faster convergence characteristics than value-based methods @cite_2 , but they usually suffer from low sample-efficiency, high variance and often converge to local optima, since they typically learn via on-policy algorithms @cite_64 @cite_35 . Even the Asynchronous Ad... | {
"cite_N": [
"@cite_35",
"@cite_64",
"@cite_60",
"@cite_53",
"@cite_42",
"@cite_65",
"@cite_45",
"@cite_2"
],
"mid": [
"1191599655",
"2119717200",
"2949608212",
"2950395671",
"2260756217",
"2121863487",
"2556958149",
"2155027007"
],
"abstract": ... |
1907.02884 | 2954108589 | Intent Detection and Slot Filling are two pillar tasks in Spoken Natural Language Understanding. Common approaches adopt joint Deep Learning architectures in attention-based recurrent frameworks. In this work, we aim at exploiting the success of "recurrence-less" models for these tasks. We introduce Bert-Joint, i.e., a... | The SF task is addressed through supervised sequence labeling approaches, e.g., MEMMs @cite_7 , CRF @cite_9 or, again, with Deep Learning, such as Recurrent Neural Networks (RNNs) @cite_4 . Deep learning research started as extensions of Deep Neural Networks and DBNs (e.g., @cite_3 ) and is sometimes merged with Condit... | {
"cite_N": [
"@cite_4",
"@cite_7",
"@cite_9",
"@cite_3",
"@cite_24",
"@cite_27",
"@cite_10"
],
"mid": [
"",
"1934019294",
"2166293310",
"2395389931",
"2094472029",
"2951008357",
"2137871902"
],
"abstract": [
"",
"Hidden Markov models (HMMs) are ... |
1812.08598 | 2905004272 | Scheduling the maintenances of nuclear power plants is a complex optimization problem, formulated in 2-stage stochastic programming for the EURO ROADEF 2010 challenge. The first level optimizes the maintenance dates and refueling decisions. The second level optimizes the production to fulfill the power demands and to e... | An open question after the Challenge was to prove dual bounds for the large size instances of the competition. (2013) furnished the first dual bounds @cite_2 using dual heuristics, the bounds proven in @cite_28 @cite_27 improved these last bounds. | {
"cite_N": [
"@cite_28",
"@cite_27",
"@cite_2"
],
"mid": [
"2567238128",
"2807125841",
"166079659"
],
"abstract": [
"",
"The EURO ROADEF 2010 Challenge aimed to schedule the maintenance and refueling operations of French nuclear power plants, ranking the approaches in competit... |
1812.08598 | 2905004272 | Scheduling the maintenances of nuclear power plants is a complex optimization problem, formulated in 2-stage stochastic programming for the EURO ROADEF 2010 challenge. The first level optimizes the maintenance dates and refueling decisions. The second level optimizes the production to fulfill the power demands and to e... | Two matheuristic approaches were designed for the ROADEF Challenge. A simplified MIP problem is solved, and the solution is repaired to ensure the feasibility for all the constraints and computing the cost in the full model. In both cases, the MIP model is solved heuristically using truncated exact methods, the simplif... | {
"cite_N": [
"@cite_6",
"@cite_8"
],
"mid": [
"2059695025",
"1583833469"
],
"abstract": [
"This paper presents a heuristic method based on column generation for the EDF (Electricite De France) long-term electricity production planning problem proposed as subject of the ROADEF EURO 2010 Ch... |
1812.08598 | 2905004272 | Scheduling the maintenances of nuclear power plants is a complex optimization problem, formulated in 2-stage stochastic programming for the EURO ROADEF 2010 challenge. The first level optimizes the maintenance dates and refueling decisions. The second level optimizes the production to fulfill the power demands and to e... | (2013) considered an exact formulation of CT6 constraints, in a CG approach dualizing coupling constraints among units @cite_6 , i.e. CT1 demands and CT14 to CT21 scheduling constraints. The MIP considered a unique scenario, the average one, with production time steps aggregated weekly. The CG approach is deployed to c... | {
"cite_N": [
"@cite_6"
],
"mid": [
"2059695025"
],
"abstract": [
"This paper presents a heuristic method based on column generation for the EDF (Electricite De France) long-term electricity production planning problem proposed as subject of the ROADEF EURO 2010 Challenge. This is to our knowledge... |
1812.08598 | 2905004272 | Scheduling the maintenances of nuclear power plants is a complex optimization problem, formulated in 2-stage stochastic programming for the EURO ROADEF 2010 challenge. The first level optimizes the maintenance dates and refueling decisions. The second level optimizes the production to fulfill the power demands and to e... | (2013) relaxed fully the constraints CT6 and CT12, leading to a MIP formulation with binaries only for the outage decisions @cite_21 . The production time steps were aggregated to weeks for size reasons. The stochastic scenarios were not aggregated, leading to 2-stage stochastic programming structure solved by Bender's... | {
"cite_N": [
"@cite_21"
],
"mid": [
"2055066666"
],
"abstract": [
"This paper describes a Benders decomposition-based framework for solving the large scale energy management problem that was posed for the ROADEF 2010 challenge. The problem was taken from the power industry and entailed scheduling... |
1812.08598 | 2905004272 | Scheduling the maintenances of nuclear power plants is a complex optimization problem, formulated in 2-stage stochastic programming for the EURO ROADEF 2010 challenge. The first level optimizes the maintenance dates and refueling decisions. The second level optimizes the production to fulfill the power demands and to e... | A natural idea to solve the problem by decomposition is to follow the 2-stage structure of stochastic programming, distinguishing the high-level maintenance and refueling problem of T2 units from the lower level production problems, as in @cite_18 @cite_19 @cite_2 @cite_23 @cite_25 @cite_3 . The high-level problem fixe... | {
"cite_N": [
"@cite_18",
"@cite_3",
"@cite_19",
"@cite_23",
"@cite_2",
"@cite_25"
],
"mid": [
"2178167101",
"2036434250",
"2501937168",
"1963524746",
"166079659",
""
],
"abstract": [
"The demand for electrical energy is globally growing very quickly. For th... |
1812.08598 | 2905004272 | Scheduling the maintenances of nuclear power plants is a complex optimization problem, formulated in 2-stage stochastic programming for the EURO ROADEF 2010 challenge. The first level optimizes the maintenance dates and refueling decisions. The second level optimizes the production to fulfill the power demands and to e... | We note that the operational approach in the French Utility Company was in this scope before the Challenge, we refer to @cite_14 . The approaches in competition were not among the most efficient, which can be analyzed comparing with the solving characteristics of frontal local search approaches. | {
"cite_N": [
"@cite_14"
],
"mid": [
"625673688"
],
"abstract": [
"Les recherches presentees dans cette these portent sur la modelisation et la resolution de systemes de contraintes, en considerant aussi bien l'aspect theorique que l'aspect pratique. La partie theorique a comme objectif de propose... |
1812.08683 | 2904933001 | In this paper, we propose a robust method to estimate the average treatment effects in observational studies when the number of potential confounders is possibly much greater than the sample size. We first use a class of penalized M-estimators for the propensity score and outcome models. We then calibrate the initial e... | In this subsection, we compare our method with the related work. First, we comment on the theoretical results of the AIPW estimator and double selection estimator when both the propensity score and outcome models are correctly specified. Second, we compare the results when one of the two models is misspecified. Finally... | {
"cite_N": [
"@cite_35",
"@cite_29",
"@cite_14",
"@cite_33"
],
"mid": [
"2261710003",
"2565113373",
"2786512041",
"2765948077"
],
"abstract": [
"In observational studies, propensity scores are commonly estimated by maxi- mum likelihood but may fail to balance high-dimensio... |
1812.08683 | 2904933001 | In this paper, we propose a robust method to estimate the average treatment effects in observational studies when the number of potential confounders is possibly much greater than the sample size. We first use a class of penalized M-estimators for the propensity score and outcome models. We then calibrate the initial e... | When both the propensity score and outcome models are correctly specified, and showed that the AIPW estimator is asymptotically normal and efficient in high dimension. Their assumptions and main results are parallel to our Theorem . However, the sample boundedness property in Remark does not hold for the AIPW estimator... | {
"cite_N": [
"@cite_18",
"@cite_8",
"@cite_21",
"@cite_32",
"@cite_23",
"@cite_10",
"@cite_25"
],
"mid": [
"2039811614",
"2137370054",
"2120817734",
"2086102241",
"1999188374",
"2058499415",
""
],
"abstract": [
"Abstract In applied problems it is co... |
1812.08683 | 2904933001 | In this paper, we propose a robust method to estimate the average treatment effects in observational studies when the number of potential confounders is possibly much greater than the sample size. We first use a class of penalized M-estimators for the propensity score and outcome models. We then calibrate the initial e... | When either the propensity score model or the outcome model is misspecified, Propositions and provide a complete characterization of the asymptotic behavior of our estimator. In the same context, @cite_3 proved that the AIPW estimator is consistent, but Theorem 2 of that work does not yield an explicit convergence rate... | {
"cite_N": [
"@cite_3"
],
"mid": [
"2100532505"
],
"abstract": [
"This paper concerns robust inference on average treatment effects following model selection. Under selection on observables, we construct confidence intervals using a doubly-robust estimator that are robust to model selection error... |
1812.08683 | 2904933001 | In this paper, we propose a robust method to estimate the average treatment effects in observational studies when the number of potential confounders is possibly much greater than the sample size. We first use a class of penalized M-estimators for the propensity score and outcome models. We then calibrate the initial e... | In another recent work, @cite_35 proposed a generalized covariate balancing method based on a class of scoring rules. Many existing covariate balancing estimators can be treated as the primal or dual problems of their optimization problem. @cite_35 studied the robustness of these estimators to misspecified propensity s... | {
"cite_N": [
"@cite_35"
],
"mid": [
"2261710003"
],
"abstract": [
"In observational studies, propensity scores are commonly estimated by maxi- mum likelihood but may fail to balance high-dimensional pre-treatment covariates even after specification search. We introduce a general framework that un... |
1812.08683 | 2904933001 | In this paper, we propose a robust method to estimate the average treatment effects in observational studies when the number of potential confounders is possibly much greater than the sample size. We first use a class of penalized M-estimators for the propensity score and outcome models. We then calibrate the initial e... | Most recently, @cite_33 @cite_14 proposed a penalized calibrated propensity score method and studied its robustness to model misspecification. Our work is closely related to @cite_33 , which can be seen as equivalent to directly plugging the initial estimator @math into the Horvitz-Thompson estimator with @math . Howev... | {
"cite_N": [
"@cite_14",
"@cite_33"
],
"mid": [
"2786512041",
"2765948077"
],
"abstract": [
"Consider the problem of estimating average treatment effects when a large number of covariates are used to adjust for possible confounding through outcome regression and propensity score models. T... |
1907.02480 | 2955607373 | The always increasing mobile connectivity affects every aspect of our daily lives, including how and when we keep ourselves informed and consult news media. By studying mobile web data, provided by one of the major Chilean telecommunication companies, we investigate how different cohorts of the population of Santiago D... | Chilean news media have been studied to outline for example their ownership structures or their political bias and to understand the effects of certain press manipulations by owners of content shown @cite_15 . These results are based on hypotheses born out of an operationalization of Herman and Chomsky's Propaganda Mod... | {
"cite_N": [
"@cite_15",
"@cite_4",
"@cite_7"
],
"mid": [
"2766427402",
"",
"2793539342"
],
"abstract": [
"CONICYT 63130228 Movistar - Telefonica Chile Chilean government initiative CORFO 13CEE2-21592 (2013-21592-1-INNOVA PRODUCCION2013-21592-1) Conicyt's Proyecto de Informacion C... |
1907.02511 | 2956103957 | In linear inverse problems, the goal is to recover a target signal from undersampled, incomplete or noisy linear measurements. Typically, the recovery relies on complex numerical optimization methods; recent approaches perform an unfolding of a numerical algorithm into a neural network form, resulting in a substantial ... | A common approach for solving problems of the form with sparsity constraints is convex optimization @cite_17 . Let us assume that the unknown @math has a sparse representation @math with respect to a dictionary @math , @math , that is, @math . Then, takes the form and a solution can be obtained via the formulation of t... | {
"cite_N": [
"@cite_17"
],
"mid": [
"2078204800"
],
"abstract": [
"The time-frequency and time-scale communities have recently developed a large number of overcomplete waveform dictionaries---stationary wavelets, wavelet packets, cosine packets, chirplets, and warplets, to name a few. Decompositi... |
1907.02511 | 2956103957 | In linear inverse problems, the goal is to recover a target signal from undersampled, incomplete or noisy linear measurements. Typically, the recovery relies on complex numerical optimization methods; recent approaches perform an unfolding of a numerical algorithm into a neural network form, resulting in a substantial ... | Numerical methods @cite_10 proposed to solve include pivoting algorithms, interior-point methods, gradient based methods and message passing algorithms (AMP) @cite_4 . Among gradient based methods, proximal methods are tailored to optimize an objective of the form where @math is a convex differentiable function with a ... | {
"cite_N": [
"@cite_13",
"@cite_4",
"@cite_28",
"@cite_0",
"@cite_31",
"@cite_10"
],
"mid": [
"",
"2166670884",
"2093545205",
"1946620893",
"2115706991",
"2118297240"
],
"abstract": [
"",
"We consider the estimation of a random vector observed through a... |
1907.02584 | 2953945697 | We propose a fast, model agnostic method for finding interpretable counterfactual explanations of classifier predictions by using class prototypes. We show that class prototypes, obtained using either an encoder or through class specific k-d trees, significantly speed up the the search for counterfactual instances and ... | The problem of local, instance level model explanations for classification can be approached from various angles. Feature attribution methods assign importance to each input feature for a given prediction. Attribution methods can be fully model agnostic @cite_18 @cite_31 or require knowledge of the architecture of the ... | {
"cite_N": [
"@cite_18",
"@cite_14",
"@cite_7",
"@cite_3",
"@cite_2",
"@cite_31",
"@cite_10",
"@cite_12"
],
"mid": [
"2282821441",
"2947285452",
"2964330603",
"2195388612",
"2773497437",
"2962862931",
"1787224781",
"2597603852"
],
"abstract": [
... |
1907.02584 | 2953945697 | We propose a fast, model agnostic method for finding interpretable counterfactual explanations of classifier predictions by using class prototypes. We show that class prototypes, obtained using either an encoder or through class specific k-d trees, significantly speed up the the search for counterfactual instances and ... | Another approach is to determine which features should remain the same so the prediction does not change. These unchanged features can be translated into called @cite_32 . Anchors are complementary to counterfactual reasoning and concepts from both approaches have been combined in the form of which consist of and @cite... | {
"cite_N": [
"@cite_4",
"@cite_22",
"@cite_9",
"@cite_21",
"@cite_1",
"@cite_32",
"@cite_16",
"@cite_34",
"@cite_13"
],
"mid": [
"2963483561",
"2099471712",
"2947794021",
"2946940672",
"",
"",
"2962760235",
"2963276306",
"2963366547"
],
... |
1907.02584 | 2953945697 | We propose a fast, model agnostic method for finding interpretable counterfactual explanations of classifier predictions by using class prototypes. We show that class prototypes, obtained using either an encoder or through class specific k-d trees, significantly speed up the the search for counterfactual instances and ... | One of the key contributions of this paper is the use of prototypes to guide the counterfactual search process. @cite_23 @cite_27 use prototypes as example-based explanations to improve the interpretability of complex datasets. Besides improving interpretability, prototypes have a broad range of applications like clust... | {
"cite_N": [
"@cite_30",
"@cite_28",
"@cite_29",
"@cite_6",
"@cite_0",
"@cite_27",
"@cite_23",
"@cite_20"
],
"mid": [
"2084701050",
"2601450892",
"2165558283",
"87092222",
"1969719708",
"2732351827",
"2551974706",
"2154642048"
],
"abstract": [
... |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.