title stringlengths 14 146 | paper_url stringlengths 105 105 | authors listlengths 0 12 | type stringclasses 0
values | primary_area stringclasses 0
values | abstract large_stringlengths 292 1.97k | keywords listlengths 0 0 | TL;DR large_stringclasses 0
values | submission_number int64 1 1.01k | arxiv_id stringlengths 10 10 ⌀ | arxiv_id_source stringclasses 2
values | embedding listlengths 768 768 |
|---|---|---|---|---|---|---|---|---|---|---|---|
Synthesized Policies for Transfer and Adaptation across Tasks and Environments | https://proceedings.neurips.cc/paper_files/paper/2018/hash/00ac8ed3b4327bdd4ebbebcb2ba10a00-Abstract.html | [
"Hexiang Hu",
"Liyu Chen",
"Boqing Gong",
"Fei Sha"
] | null | null | The ability to transfer in reinforcement learning is key towards building an agent of general artificial intelligence. In this paper, we consider the problem of learning to simultaneously transfer across both environments and tasks, probably more importantly, by learning from only sparse (environment, task) pairs out o... | [] | null | 1 | 1904.03276 | title_snapshot | [
0.0012224720558151603,
-0.020996607840061188,
0.006565372925251722,
0.03466087579727173,
0.04165482893586159,
0.029941676184535027,
-0.005667293444275856,
-0.0054527269676327705,
-0.02362792007625103,
-0.03228210285305977,
-0.009545942768454552,
0.01585470698773861,
-0.07860804349184036,
-... |
Self-Supervised Generation of Spatial Audio for 360° Video | https://proceedings.neurips.cc/paper_files/paper/2018/hash/01161aaa0b6d1345dd8fe4e481144d84-Abstract.html | [
"Pedro Morgado",
"Nuno Nvasconcelos",
"Timothy Langlois",
"Oliver Wang"
] | null | null | We introduce an approach to convert mono audio recorded by a 360° video camera into spatial audio, a representation of the distribution of sound over the full viewing sphere. Spatial audio is an important component of immersive 360° video viewing, but spatial audio microphones are still rare in current 360° video produ... | [] | null | 2 | 1809.02587 | title_snapshot | [
0.04828796908259392,
0.02129501849412918,
0.011184366419911385,
0.04540909081697464,
0.0332086905837059,
0.04181196913123131,
0.03995216637849808,
0.0012549204984679818,
-0.022972093895077705,
-0.06532706320285797,
-0.014347396790981293,
-0.006411992944777012,
-0.05857205390930176,
0.03209... |
On GANs and GMMs | https://proceedings.neurips.cc/paper_files/paper/2018/hash/0172d289da48c48de8c5ebf3de9f7ee1-Abstract.html | [
"Eitan Richardson",
"Yair Weiss"
] | null | null | A longstanding problem in machine learning is to find unsupervised methods that can learn the statistical structure of high dimensional signals. In recent years, GANs have gained much attention as a possible solution to the problem, and in particular have shown the ability to generate remarkably realistic high resoluti... | [] | null | 3 | 1805.12462 | title_snapshot | [
-0.016028378158807755,
-0.02084904909133911,
-0.010864353738725185,
0.04551743343472481,
0.02786976285278797,
0.024190135300159454,
0.022184127941727638,
0.023071618750691414,
-0.028016144409775734,
-0.050372231751680374,
-0.01998254284262657,
0.022238630801439285,
-0.06219659000635147,
0.... |
Batch-Instance Normalization for Adaptively Style-Invariant Neural Networks | https://proceedings.neurips.cc/paper_files/paper/2018/hash/018b59ce1fd616d874afad0f44ba338d-Abstract.html | [
"Hyeonseob Nam",
"Hyo-Eun Kim"
] | null | null | Real-world image recognition is often challenged by the variability of visual styles including object textures, lighting conditions, filter effects, etc. Although these variations have been deemed to be implicitly handled by more training data and deeper networks, recent advances in image style transfer suggest that it... | [] | null | 4 | 1805.07925 | title_snapshot | [
0.012399038299918175,
-0.014829735271632671,
-0.005317362956702709,
0.030226614326238632,
0.03253081440925598,
0.056409820914268494,
0.000837739382404834,
0.007866326719522476,
-0.02802116982638836,
-0.05529319494962692,
-0.03974480554461479,
-0.012440280988812447,
-0.06646210700273514,
-0... |
Hierarchical Reinforcement Learning for Zero-shot Generalization with Subtask Dependencies | https://proceedings.neurips.cc/paper_files/paper/2018/hash/018dd1e07a2de4a08e6612341bf2323e-Abstract.html | [
"Sungryull Sohn",
"Junhyuk Oh",
"Honglak Lee"
] | null | null | We introduce a new RL problem where the agent is required to generalize to a previously-unseen environment characterized by a subtask graph which describes a set of subtasks and their dependencies. Unlike existing hierarchical multitask RL approaches that explicitly describe what the agent should do at a high level, ou... | [] | null | 5 | 1807.07665 | title_snapshot | [
-0.02905353531241417,
-0.02951153554022312,
0.02911578118801117,
0.04968457669019699,
0.04343939945101738,
0.03453923761844635,
0.03306092694401741,
0.000017163380107376724,
-0.03418993949890137,
-0.038721416145563126,
-0.014685881324112415,
0.029560303315520287,
-0.07674241065979004,
-0.0... |
KDGAN: Knowledge Distillation with Generative Adversarial Networks | https://proceedings.neurips.cc/paper_files/paper/2018/hash/019d385eb67632a7e958e23f24bd07d7-Abstract.html | [
"Xiaojie Wang",
"Rui Zhang",
"Yu Sun",
"Jianzhong Qi"
] | null | null | Knowledge distillation (KD) aims to train a lightweight classifier suitable to provide accurate inference with constrained resources in multi-label learning. Instead of directly consuming feature-label pairs, the classifier is trained by a teacher, i.e., a high-capacity model whose training may be resource-hungry. The ... | [] | null | 6 | null | null | [
-0.01085592620074749,
-0.014297455549240112,
-0.024495797231793404,
0.0639876052737236,
0.02536129392683506,
-0.007963165640830994,
0.014924970455467701,
-0.026075441390275955,
0.0028766398318111897,
-0.024608217179775238,
-0.028947437182068825,
0.012033313512802124,
-0.06566260010004044,
... |
Contour location via entropy reduction leveraging multiple information sources | https://proceedings.neurips.cc/paper_files/paper/2018/hash/01a0683665f38d8e5e567b3b15ca98bf-Abstract.html | [
"Alexandre Marques",
"Remi Lam",
"Karen Willcox"
] | null | null | We introduce an algorithm to locate contours of functions that are expensive to evaluate. The problem of locating contours arises in many applications, including classification, constrained optimization, and performance analysis of mechanical and dynamical systems (reliability, probability of failure, stability, etc.).... | [] | null | 7 | 1805.07489 | title_snapshot | [
-0.01301408652216196,
0.009537380188703537,
-0.02070181630551815,
0.04668501392006874,
0.05638739466667175,
0.07139978557825089,
0.006688088644295931,
-0.016874561086297035,
-0.042548272758722305,
-0.07688795030117035,
-0.020190779119729996,
0.005815932061523199,
-0.035107795149087906,
-0.... |
Contextual bandits with surrogate losses: Margin bounds and efficient algorithms | https://proceedings.neurips.cc/paper_files/paper/2018/hash/01e9565cecc4e989123f9620c1d09c09-Abstract.html | [
"Dylan J Foster",
"Akshay Krishnamurthy"
] | null | null | We use surrogate losses to obtain several new regret bounds and new algorithms for contextual bandit learning. Using the ramp loss, we derive a new margin-based regret bound in terms of standard sequential complexity measures of a benchmark class of real-valued regression functions. Using the hinge loss, we derive an e... | [] | null | 8 | 1806.10745 | title_snapshot | [
-0.03195571526885033,
0.007576164323836565,
-0.000709768442902714,
0.04422660544514656,
0.02683165855705738,
0.023374047130346298,
0.03484922647476196,
-0.007824786007404327,
-0.03125830367207527,
-0.0359438881278038,
-0.0036579479929059744,
0.025831764563918114,
-0.05906963720917702,
-0.0... |
Adaptive Sampling Towards Fast Graph Representation Learning | https://proceedings.neurips.cc/paper_files/paper/2018/hash/01eee509ee2f68dc6014898c309e86bf-Abstract.html | [
"Wenbing Huang",
"Tong Zhang",
"Yu Rong",
"Junzhou Huang"
] | null | null | Graph Convolutional Networks (GCNs) have become a crucial tool on learning representations of graph vertices. The main challenge of adapting GCNs on large-scale graphs is the scalability issue that it incurs heavy cost both in computation and memory due to the uncontrollable neighborhood expansion across layers. In thi... | [] | null | 9 | 1809.05343 | title_snapshot | [
-0.0021381410770118237,
-0.05493427813053131,
0.01032721996307373,
0.04996773228049278,
0.03729455918073654,
0.029517140239477158,
0.009885689243674278,
0.02523386850953102,
-0.005518842488527298,
-0.0716748982667923,
0.022390881553292274,
-0.0474889911711216,
-0.08690950274467468,
0.02281... |
Analytic solution and stationary phase approximation for the Bayesian lasso and elastic net | https://proceedings.neurips.cc/paper_files/paper/2018/hash/0245952ecff55018e2a459517fdb40e3-Abstract.html | [
"Tom Michoel"
] | null | null | The lasso and elastic net linear regression models impose a double-exponential prior distribution on the model parameters to achieve regression shrinkage and variable selection, allowing the inference of robust models from large data sets. However, there has been limited success in deriving estimates for the full poste... | [] | null | 10 | 1709.08535 | title_snapshot | [
-0.014020564034581184,
-0.01324667502194643,
0.006448504514992237,
-0.011159011162817478,
0.050265729427337646,
0.05149402841925621,
0.027327097952365875,
-0.018893208354711533,
-0.054496899247169495,
-0.06366017460823059,
0.020754283294081688,
-0.00335368188098073,
-0.06683680415153503,
-... |
Identification and Estimation of Causal Effects from Dependent Data | https://proceedings.neurips.cc/paper_files/paper/2018/hash/024677efb8e4aee2eaeef17b54695bbe-Abstract.html | [
"Eli Sherman",
"Ilya Shpitser"
] | null | null | The assumption that data samples are independent and identically distributed (iid) is standard in many areas of statistics and machine learning. Nevertheless, in some settings, such as social networks, infectious disease modeling, and reasoning with spatial and temporal data, this assumption is false. An extensive lite... | [] | null | 11 | 1902.01443 | title_snapshot | [
0.027190767228603363,
-0.009279330261051655,
-0.02883763238787651,
0.03284650668501854,
0.04497356712818146,
0.03394180163741112,
0.05087944120168686,
0.004876905586570501,
-0.01980939321219921,
-0.022372275590896606,
0.02751675806939602,
-0.01595951057970524,
-0.07745250314474106,
-0.0091... |
Streamlining Variational Inference for Constraint Satisfaction Problems | https://proceedings.neurips.cc/paper_files/paper/2018/hash/02ed812220b0705fabb868ddbf17ea20-Abstract.html | [
"Aditya Grover",
"Tudor Achim",
"Stefano Ermon"
] | null | null | Several algorithms for solving constraint satisfaction problems are based on survey propagation, a variational inference scheme used to obtain approximate marginal probability estimates for variable assignments. These marginals correspond to how frequently each variable is set to true among satisfying assignments, and ... | [] | null | 12 | 1811.09813 | title_snapshot | [
-0.02877364307641983,
-0.00023990495537873358,
-0.01234692707657814,
0.06088383495807648,
0.05146612972021103,
0.05093022808432579,
0.032837167382240295,
-0.003269853536039591,
-0.03217139467597008,
-0.009183607064187527,
-0.037361226975917816,
0.024291131645441055,
-0.0705905333161354,
0.... |
A Spectral View of Adversarially Robust Features | https://proceedings.neurips.cc/paper_files/paper/2018/hash/033cc385728c51d97360020ed57776f0-Abstract.html | [
"Shivam Garg",
"Vatsal Sharan",
"Brian Zhang",
"Gregory Valiant"
] | null | null | Given the apparent difficulty of learning models that are robust to adversarial perturbations, we propose tackling the simpler problem of developing adversarially robust features. Specifically, given a dataset and metric of interest, the goal is to return a function (or multiple functions) that 1) is robust to adversar... | [] | null | 13 | 1811.06609 | title_snapshot | [
-0.007039261050522327,
-0.029013652354478836,
-0.01193443313241005,
0.04721450433135033,
0.035524819046258926,
0.01899353601038456,
0.04395038262009621,
-0.042900118976831436,
-0.011618237942457199,
-0.062180936336517334,
-0.037592727690935135,
0.0023418942000716925,
-0.07701505720615387,
... |
Dimensionality Reduction has Quantifiable Imperfections: Two Geometric Bounds | https://proceedings.neurips.cc/paper_files/paper/2018/hash/037a595e6f4f0576a9efe43154d71c18-Abstract.html | [
"Kry Lui",
"Gavin Weiguang Ding",
"Ruitong Huang",
"Robert McCann"
] | null | null | In this paper, we investigate Dimensionality reduction (DR) maps in an information retrieval setting from a quantitative topology point of view. In particular, we show that no DR maps can achieve perfect precision and perfect recall simultaneously. Thus a continuous DR map must have imperfect precision. We further prov... | [] | null | 14 | 1811.00115 | title_snapshot | [
-0.0630987286567688,
-0.012167797423899174,
-0.0059023890644311905,
0.05886099115014076,
0.02570137195289135,
0.026129048317670822,
0.03314756974577904,
-0.00972925778478384,
-0.013661859557032585,
-0.055901478976011276,
-0.029046084731817245,
-0.031159093603491783,
-0.05479098856449127,
0... |
Learning SMaLL Predictors | https://proceedings.neurips.cc/paper_files/paper/2018/hash/03b2ceb73723f8b53cd533e4fba898ee-Abstract.html | [
"Vikas Garg",
"Ofer Dekel",
"Lin Xiao"
] | null | null | We introduce a new framework for learning in severely resource-constrained settings. Our technique delicately amalgamates the representational richness of multiple linear predictors with the sparsity of Boolean relaxations, and thereby yields classifiers that are compact, interpretable, and accurate. We provide a rigor... | [] | null | 15 | 1803.02388 | title_snapshot | [
-0.022642172873020172,
-0.009425490163266659,
0.0018342704279348254,
0.04148487746715546,
0.041001468896865845,
0.055856820195913315,
0.007071639411151409,
-0.023503335192799568,
-0.059540122747421265,
-0.001433607772924006,
-0.003077942878007889,
0.01706118881702423,
-0.07222114503383636,
... |
ResNet with one-neuron hidden layers is a Universal Approximator | https://proceedings.neurips.cc/paper_files/paper/2018/hash/03bfc1d4783966c69cc6aef8247e0103-Abstract.html | [
"Hongzhou Lin",
"Stefanie Jegelka"
] | null | null | We demonstrate that a very deep ResNet with stacked modules that have one neuron per hidden layer and ReLU activation functions can uniformly approximate any Lebesgue integrable function in d dimensions, i.e. \ell_1(R^d). Due to the identity mapping inherent to ResNets, our network has alternating layers of dimension o... | [] | null | 16 | 1806.10909 | title_snapshot | [
-0.02358231134712696,
-0.024879848584532738,
0.006486149970442057,
0.020267922431230545,
0.04285725578665733,
0.045091718435287476,
0.04050478711724281,
-0.0288814939558506,
-0.04418224096298218,
-0.036070726811885834,
0.007904107682406902,
-0.03269870951771736,
-0.0705915316939354,
-0.006... |
How Many Samples are Needed to Estimate a Convolutional Neural Network? | https://proceedings.neurips.cc/paper_files/paper/2018/hash/03c6b06952c750899bb03d998e631860-Abstract.html | [
"Simon S Du",
"Yining Wang",
"Xiyu Zhai",
"Sivaraman Balakrishnan",
"Ruslan Salakhutdinov",
"Aarti Singh"
] | null | null | A widespread folklore for explaining the success of Convolutional Neural Networks (CNNs) is that CNNs use a more compact representation than the Fully-connected Neural Network (FNN) and thus require fewer training samples to accurately estimate their parameters. We initiate the study of rigorously characterizing the sa... | [] | null | 17 | 1805.07883 | title_judge | [
0.0023242183960974216,
-0.032294780015945435,
0.0034562265500426292,
0.03380723297595978,
0.03922651708126068,
0.03366513177752495,
0.022180192172527313,
0.01817101053893566,
-0.030718734487891197,
-0.039193104952573776,
0.018663575872778893,
-0.029869547113776207,
-0.07842031866312027,
-0... |
Objective and efficient inference for couplings in neuronal networks | https://proceedings.neurips.cc/paper_files/paper/2018/hash/03cf87174debaccd689c90c34577b82f-Abstract.html | [
"Yu Terada",
"Tomoyuki Obuchi",
"Takuya Isomura",
"Yoshiyuki Kabashima"
] | null | null | Inferring directional couplings from the spike data of networks is desired in various scientific fields such as neuroscience. Here, we apply a recently proposed objective procedure to the spike data obtained from the Hodgkin-Huxley type models and in vitro neuronal networks cultured in a circular structure. As a result... | [] | null | 18 | 1805.07061 | title_snapshot | [
-0.049758344888687134,
0.026216277852654457,
0.006426689680665731,
0.007289254572242498,
0.03078841231763363,
0.01330373715609312,
0.0469038151204586,
0.000980343553237617,
-0.043181829154491425,
-0.037581462413072586,
0.03956517577171326,
-0.0010343893663957715,
-0.07420965284109116,
0.00... |
Unsupervised Adversarial Invariance | https://proceedings.neurips.cc/paper_files/paper/2018/hash/03e7ef47cee6fa4ae7567394b99912b7-Abstract.html | [
"Ayush Jaiswal",
"Rex Yue Wu",
"Wael Abd-Almageed",
"Prem Natarajan"
] | null | null | Data representations that contain all the information about target variables but are invariant to nuisance factors benefit supervised learning algorithms by preventing them from learning associations between these factors and the targets, thus reducing overfitting. We present a novel unsupervised invariance induction f... | [] | null | 19 | 1809.10083 | title_snapshot | [
0.010314730927348137,
-0.0385354720056057,
-0.011103500612080097,
0.02130839042365551,
0.02708439715206623,
0.004830105230212212,
0.04234371334314346,
-0.028506316244602203,
-0.03570452705025673,
-0.012656143866479397,
-0.02120582014322281,
0.006584309507161379,
-0.08844815194606781,
0.003... |
Critical initialisation for deep signal propagation in noisy rectifier neural networks | https://proceedings.neurips.cc/paper_files/paper/2018/hash/045cf83ab0722e782cf72d14e44adf98-Abstract.html | [
"Arnu Pretorius",
"Elan van Biljon",
"Steve Kroon",
"Herman Kamper"
] | null | null | Stochastic regularisation is an important weapon in the arsenal of a deep learning practitioner. However, despite recent theoretical advances, our understanding of how noise influences signal propagation in deep neural networks remains limited. By extending recent work based on mean field theory, we develop a new frame... | [] | null | 20 | 1811.00293 | title_snapshot | [
-0.010525071993470192,
-0.019101016223430634,
0.01288637612015009,
0.00889857392758131,
0.019698122516274452,
0.052002519369125366,
0.029611807316541672,
-0.0014516571536660194,
-0.03724382817745209,
-0.049228768795728683,
0.012024415656924248,
0.013957316055893898,
-0.05086739733815193,
0... |
Learning sparse neural networks via sensitivity-driven regularization | https://proceedings.neurips.cc/paper_files/paper/2018/hash/04df4d434d481c5bb723be1b6df1ee65-Abstract.html | [
"Enzo Tartaglione",
"Skjalg Lepsøy",
"Attilio Fiandrotti",
"Gianluca Francini"
] | null | null | The ever-increasing number of parameters in deep neural networks poses challenges for memory-limited applications. Regularize-and-prune methods aim at meeting these challenges by sparsifying the network weights. In this context we quantify the output sensitivity to the parameters (i.e. their relevance to the network ou... | [] | null | 21 | 1810.11764 | title_snapshot | [
-0.032429855316877365,
-0.012893510982394218,
0.00755516579374671,
0.017795683816075325,
0.044992756098508835,
0.06772710382938385,
0.013493579812347889,
-0.021038997918367386,
-0.04412325471639633,
-0.022342531010508537,
0.0005749005940742791,
0.005296891089528799,
-0.04991660639643669,
0... |
Cooperative neural networks (CoNN): Exploiting prior independence structure for improved classification | https://proceedings.neurips.cc/paper_files/paper/2018/hash/051928341be67dcba03f0e04104d9047-Abstract.html | [
"Harsh Shrivastava",
"Eugene Bart",
"Bob Price",
"Hanjun Dai",
"Bo Dai",
"Srinivas Aluru"
] | null | null | We propose a new approach, called cooperative neural networks (CoNN), which use a set of cooperatively trained neural networks to capture latent representations that exploit prior given independence structure. The model is more flexible than traditional graphical models based on exponential family distributions, but in... | [] | null | 22 | 1906.00291 | title_snapshot | [
-0.028974266722798347,
-0.0193710345774889,
-0.026008784770965576,
0.04869880899786949,
0.03792363405227661,
0.0365142859518528,
0.002024570945650339,
-0.005300480406731367,
-0.01836145482957363,
-0.022701626643538475,
0.002265286399051547,
0.014748580753803253,
-0.06581184267997742,
0.011... |
Data center cooling using model-predictive control | https://proceedings.neurips.cc/paper_files/paper/2018/hash/059fdcd96baeb75112f09fa1dcc740cc-Abstract.html | [
"Nevena Lazic",
"Craig Boutilier",
"Tyler Lu",
"Eehern Wong",
"Binz Roy",
"MK Ryu",
"Greg Imwalle"
] | null | null | Despite impressive recent advances in reinforcement learning (RL), its deployment in real-world physical systems is often complicated by unexpected events, limited data, and the potential for expensive failures. In this paper, we describe an application of RL “in the wild” to the task of regulating temperatures and air... | [] | null | 23 | null | null | [
-0.021946867927908897,
-0.043773550540208817,
-0.009827534668147564,
0.04511412978172302,
0.08032295852899551,
0.01076729316264391,
0.016462190076708794,
-0.0037919541355222464,
-0.02976994961500168,
-0.0029145677108317614,
-0.036021459847688675,
-0.024261120706796646,
-0.07129943370819092,
... |
Generalization Bounds for Uniformly Stable Algorithms | https://proceedings.neurips.cc/paper_files/paper/2018/hash/05a624166c8eb8273b8464e8d9cb5bd9-Abstract.html | [
"Vitaly Feldman",
"Jan Vondrak"
] | null | null | Uniform stability of a learning algorithm is a classical notion of algorithmic stability introduced to derive high-probability bounds on the generalization error (Bousquet and Elisseeff, 2002). Specifically, for a loss function with range bounded in $[0,1]$, the generalization error of $\gamma$-uniformly stable learnin... | [] | null | 24 | 1812.09859 | title_snapshot | [
-0.018344419077038765,
-0.006676559802144766,
-0.018095925450325012,
0.03243451192975044,
0.042303308844566345,
0.016754550859332085,
0.03157774358987808,
-0.02303934283554554,
-0.02255929633975029,
-0.037201955914497375,
-0.002362212399020791,
-0.012783441692590714,
-0.09613209962844849,
... |
COLA: Decentralized Linear Learning | https://proceedings.neurips.cc/paper_files/paper/2018/hash/05a70454516ecd9194c293b0e415777f-Abstract.html | [
"Lie He",
"An Bian",
"Martin Jaggi"
] | null | null | Decentralized machine learning is a promising emerging paradigm in view of global challenges of data ownership and privacy. We consider learning of linear classification and regression models, in the setting where the training data is decentralized over many user devices, and the learning algorithm must run on-device, ... | [] | null | 25 | 1808.04883 | title_snapshot | [
0.0023041851818561554,
-0.02875504270195961,
-0.017417922616004944,
0.03171030431985855,
0.035434577614068985,
0.03004412166774273,
0.020895307883620262,
-0.01504290197044611,
-0.0076002501882612705,
-0.03150194138288498,
0.00601218780502677,
-0.006301854271441698,
-0.08768540620803833,
0.... |
Leveraging the Exact Likelihood of Deep Latent Variable Models | https://proceedings.neurips.cc/paper_files/paper/2018/hash/0609154fa35b3194026346c9cac2a248-Abstract.html | [
"Pierre-Alexandre Mattei",
"Jes Frellsen"
] | null | null | Deep latent variable models (DLVMs) combine the approximation abilities of deep neural networks and the statistical foundations of generative models. Variational methods are commonly used for inference; however, the exact likelihood of these models has been largely overlooked. The purpose of this work is to study the g... | [] | null | 26 | 1802.04826 | title_snapshot | [
0.0018646763637661934,
-0.025026876479387283,
-0.02707708068192005,
0.05339581519365311,
0.047532204538583755,
0.05421138182282448,
0.03816356137394905,
0.01857130229473114,
-0.03718511015176773,
-0.043510545045137405,
-0.01577906310558319,
0.012346392497420311,
-0.049686405807733536,
0.02... |
A General Method for Amortizing Variational Filtering | https://proceedings.neurips.cc/paper_files/paper/2018/hash/060afc8a563aaccd288f98b7c8723b61-Abstract.html | [
"Joseph Marino",
"Milan Cvitkovic",
"Yisong Yue"
] | null | null | We introduce the variational filtering EM algorithm, a simple, general-purpose method for performing variational inference in dynamical latent variable models using information from only past and present variables, i.e. filtering. The algorithm is derived from the variational objective in the filtering setting and cons... | [] | null | 27 | 1811.05090 | title_snapshot | [
0.00449665542691946,
0.022106952965259552,
-0.007216393016278744,
0.04562665522098541,
0.04331256449222565,
0.03374956175684929,
0.040446847677230835,
0.01389323826879263,
-0.056407153606414795,
-0.03102760575711727,
0.004775531589984894,
0.021437492221593857,
-0.07638412714004517,
0.00491... |
One-Shot Unsupervised Cross Domain Translation | https://proceedings.neurips.cc/paper_files/paper/2018/hash/062ddb6c727310e76b6200b7c71f63b5-Abstract.html | [
"Sagie Benaim",
"Lior Wolf"
] | null | null | Given a single image $x$ from domain $A$ and a set of images from domain $B$, our task is to generate the analogous of $x$ in $B$. We argue that this task could be a key AI capability that underlines the ability of cognitive agents to act in the world and present empirical evidence that the existing unsupervised domain... | [] | null | 28 | 1806.06029 | title_snapshot | [
0.013726992532610893,
-0.018645361065864563,
-0.014419473707675934,
0.045952387154102325,
0.03400539606809616,
0.022025112062692642,
0.030397245660424232,
0.0007953173480927944,
0.013721995055675507,
-0.03633050620555878,
-0.03255157917737961,
0.00944045465439558,
-0.06842249631881714,
-0.... |
Query K-means Clustering and the Double Dixie Cup Problem | https://proceedings.neurips.cc/paper_files/paper/2018/hash/0655f117444fc1911ab9c6f6b0139051-Abstract.html | [
"I Chien",
"Chao Pan",
"Olgica Milenkovic"
] | null | null | We consider the problem of approximate $K$-means clustering with outliers and side information provided by same-cluster queries and possibly noisy answers. Our solution shows that, under some mild assumptions on the smallest cluster size, one can obtain an $(1+\epsilon)$-approximation for the optimal potential with pro... | [] | null | 29 | 1806.05938 | title_snapshot | [
-0.017954275012016296,
0.002261081477627158,
0.01097283698618412,
0.037891749292612076,
0.05175592377781868,
0.03871412202715874,
0.03219744935631752,
-0.0017736790468916297,
-0.01881149783730507,
-0.043864656239748,
-0.02082478441298008,
-0.027908729389309883,
-0.051257964223623276,
0.021... |
Probabilistic Neural Programmed Networks for Scene Generation | https://proceedings.neurips.cc/paper_files/paper/2018/hash/06964dce9addb1c5cb5d6e3d9838f733-Abstract.html | [
"Zhiwei Deng",
"Jiacheng Chen",
"YIFANG FU",
"Greg Mori"
] | null | null | In this paper we address the text to scene image generation problem. Generative models that capture the variability in complicated scenes containing rich semantics is a grand goal of image generation. Complicated scene images contain rich visual elements, compositional visual concepts, and complicated relations between... | [] | null | 30 | null | null | [
0.01740226522088051,
0.0027882708236575127,
-0.01287858933210373,
0.06247006729245186,
0.0415610708296299,
0.025433629751205444,
0.0015532942488789558,
0.022817173972725868,
-0.026983126997947693,
-0.039465755224227905,
-0.041892051696777344,
-0.025767510756850243,
-0.06374605745077133,
-0... |
Escaping Saddle Points in Constrained Optimization | https://proceedings.neurips.cc/paper_files/paper/2018/hash/069654d5ce089c13f642d19f09a3d1c0-Abstract.html | [
"Aryan Mokhtari",
"Asuman Ozdaglar",
"Ali Jadbabaie"
] | null | null | In this paper, we study the problem of escaping from saddle points in smooth nonconvex optimization problems subject to a convex set $\mathcal{C}$. We propose a generic framework that yields convergence to a second-order stationary point of the problem, if the convex set $\mathcal{C}$ is simple for a quadratic objectiv... | [] | null | 31 | 1809.02162 | title_snapshot | [
-0.04991958290338516,
-0.02373429946601391,
0.004711022600531578,
0.045168686658144,
0.021868310868740082,
0.039792440831661224,
0.02848568558692932,
0.007978684268891811,
-0.04297453910112381,
-0.05105448141694069,
-0.02199861966073513,
-0.008288230746984482,
-0.055807940661907196,
-0.013... |
Adversarial Text Generation via Feature-Mover's Distance | https://proceedings.neurips.cc/paper_files/paper/2018/hash/074177d3eb6371e32c16c55a3b8f706b-Abstract.html | [
"Liqun Chen",
"Shuyang Dai",
"Chenyang Tao",
"Haichao Zhang",
"Zhe Gan",
"Dinghan Shen",
"Yizhe Zhang",
"Guoyin Wang",
"Ruiyi Zhang",
"Lawrence Carin"
] | null | null | Generative adversarial networks (GANs) have achieved significant success in generating real-valued data. However, the discrete nature of text hinders the application of GAN to text-generation tasks. Instead of using the standard GAN objective, we propose to improve text-generation GAN via a novel approach inspired by o... | [] | null | 32 | 1809.06297 | title_snapshot | [
-0.04619388282299042,
-0.02628326416015625,
-0.023502124473452568,
0.06774459779262543,
0.030678680166602135,
0.013394907116889954,
0.024991575628519058,
0.005686335731297731,
0.015697253867983818,
-0.04138605669140816,
-0.0211323332041502,
0.012036307714879513,
-0.06149831786751747,
0.024... |
On gradient regularizers for MMD GANs | https://proceedings.neurips.cc/paper_files/paper/2018/hash/07f75d9144912970de5a09f5a305e10c-Abstract.html | [
"Michael Arbel",
"Danica J. Sutherland",
"Mikołaj Bińkowski",
"Arthur Gretton"
] | null | null | We propose a principled method for gradient-based regularization of the critic of GAN-like models trained by adversarially optimizing the kernel of a Maximum Mean Discrepancy (MMD). We show that controlling the gradient of the critic is vital to having a sensible loss function, and devise a method to enforce exact, ana... | [] | null | 33 | 1805.11565 | title_snapshot | [
-0.014037431217730045,
-0.010482117533683777,
0.0002913321659434587,
0.03472789376974106,
0.027982357889413834,
0.02170710265636444,
0.012889713048934937,
-0.009621595032513142,
-0.04135299474000931,
-0.04683483764529228,
-0.023215197026729584,
0.02026389352977276,
-0.05215703696012497,
0.... |
Differentially Private Bayesian Inference for Exponential Families | https://proceedings.neurips.cc/paper_files/paper/2018/hash/08040837089cdf46631a10aca5258e16-Abstract.html | [
"Garrett Bernstein",
"Daniel R. Sheldon"
] | null | null | The study of private inference has been sparked by growing concern regarding the analysis of data when it stems from sensitive sources. We present the first method for private Bayesian inference in exponential families that properly accounts for noise introduced by the privacy mechanism. It is efficient because it work... | [] | null | 34 | 1809.02188 | title_snapshot | [
0.0028641032986342907,
0.035698842257261276,
-0.004804864525794983,
0.02996828407049179,
0.06728428602218628,
0.035300955176353455,
0.04640461504459381,
-0.0401904471218586,
-0.0136495940387249,
-0.05021539330482483,
0.01721220836043358,
0.018466489389538765,
-0.04877762869000435,
-0.01054... |
Stochastic Primal-Dual Method for Empirical Risk Minimization with O(1) Per-Iteration Complexity | https://proceedings.neurips.cc/paper_files/paper/2018/hash/08048a9c5630ccb67789a198f35d30ec-Abstract.html | [
"Conghui Tan",
"Tong Zhang",
"Shiqian Ma",
"Ji Liu"
] | null | null | Regularized empirical risk minimization problem with linear predictor appears frequently in machine learning. In this paper, we propose a new stochastic primal-dual method to solve this class of problems. Different from existing methods, our proposed methods only require O(1) operations in each iteration. We also devel... | [] | null | 35 | 1811.01182 | title_snapshot | [
-0.022922834381461143,
-0.0018158081220462918,
0.004194832406938076,
0.034376613795757294,
0.03191527724266052,
0.07132773846387863,
0.022202270105481148,
-0.01083296537399292,
-0.022792167961597443,
-0.05344100296497345,
-0.009087659418582916,
-0.013641642406582832,
-0.0497448705136776,
-... |
Answerer in Questioner's Mind: Information Theoretic Approach to Goal-Oriented Visual Dialog | https://proceedings.neurips.cc/paper_files/paper/2018/hash/0829424ffa0d3a2547b6c9622c77de03-Abstract.html | [
"Sang-Woo Lee",
"Yu-Jung Heo",
"Byoung-Tak Zhang"
] | null | null | Goal-oriented dialog has been given attention due to its numerous applications in artificial intelligence. Goal-oriented dialogue tasks occur when a questioner asks an action-oriented question and an answerer responds with the intent of letting the questioner know a correct action to take. To ask the adequate question,... | [] | null | 36 | 1802.03881 | title_snapshot | [
-0.01629101112484932,
-0.028785469010472298,
-0.0032922946847975254,
-0.002368397545069456,
0.023043645545840263,
0.03009635955095291,
0.01492038182914257,
0.00046162065700627863,
-0.03398662805557251,
-0.008427918888628483,
-0.04526127129793167,
0.0338219590485096,
-0.054909173399209976,
... |
Learning Plannable Representations with Causal InfoGAN | https://proceedings.neurips.cc/paper_files/paper/2018/hash/08aac6ac98e59e523995c161e57875f5-Abstract.html | [
"Thanard Kurutach",
"Aviv Tamar",
"Ge Yang",
"Stuart Russell",
"Pieter Abbeel"
] | null | null | In recent years, deep generative models have been shown to 'imagine' convincing high-dimensional observations such as images, audio, and even video, learning directly from raw data. In this work, we ask how to imagine goal-directed visual plans -- a plausible sequence of observations that transition a dynamical system ... | [] | null | 37 | 1807.09341 | title_snapshot | [
-0.020457468926906586,
-0.02257620543241501,
-0.04223112761974335,
0.034547142684459686,
0.021641282364726067,
0.009111630730330944,
0.008917951956391335,
-0.0018808263121172786,
-0.031798139214515686,
-0.03279269114136696,
-0.020398063585162163,
0.00002356324330321513,
-0.060662850737571716... |
Interactive Structure Learning with Structural Query-by-Committee | https://proceedings.neurips.cc/paper_files/paper/2018/hash/08c5433a60135c32e34f46a71175850c-Abstract.html | [
"Christopher Tosh",
"Sanjoy Dasgupta"
] | null | null | In this work, we introduce interactive structure learning, a framework that unifies many different interactive learning tasks. We present a generalization of the query-by-committee active learning algorithm for this setting, and we study its consistency and rate of convergence, both theoretically and empirically, with ... | [] | null | 38 | null | null | [
0.0054480344988405704,
-0.02268807590007782,
-0.007078879978507757,
0.025096260011196136,
0.01829667016863823,
-0.009793615899980068,
0.014881755225360394,
-0.014645181596279144,
-0.0229935385286808,
-0.020738326013088226,
-0.009823824279010296,
0.002034817123785615,
-0.07060176134109497,
... |
The streaming rollout of deep networks - towards fully model-parallel execution | https://proceedings.neurips.cc/paper_files/paper/2018/hash/08f90c1a417155361a5c4b8d297e0d78-Abstract.html | [
"Volker Fischer",
"Jan Koehler",
"Thomas Pfeil"
] | null | null | Deep neural networks, and in particular recurrent networks, are promising candidates to control autonomous agents that interact in real-time with the physical world. However, this requires a seamless integration of temporal features into the network’s architecture. For the training of and inference with recurrent neura... | [] | null | 39 | 1806.04965 | title_snapshot | [
-0.008214700035750866,
-0.04557248204946518,
-0.0044558788649737835,
0.04365752637386322,
0.04319709166884422,
0.025232169777154922,
0.0030443959403783083,
0.03754496946930885,
-0.010570885613560677,
-0.030561748892068863,
-0.023875080049037933,
-0.023138128221035004,
-0.06613576412200928,
... |
Contextual Stochastic Block Models | https://proceedings.neurips.cc/paper_files/paper/2018/hash/08fc80de8121419136e443a70489c123-Abstract.html | [
"Yash Deshpande",
"Subhabrata Sen",
"Andrea Montanari",
"Elchanan Mossel"
] | null | null | We provide the first information theoretical tight analysis for inference of latent community structure given a sparse graph along with high dimensional node covariates, correlated with the same latent communities. Our work bridges recent theoretical breakthroughs in detection of latent community structure without node... | [] | null | 40 | 1807.09596 | title_snapshot | [
0.029854722321033478,
0.0006911833770573139,
-0.03180632367730141,
0.03989057615399361,
0.03580968827009201,
0.029421571642160416,
0.043704040348529816,
0.010139054618775845,
-0.034662965685129166,
-0.01545302476733923,
0.01288675982505083,
-0.024402683600783348,
-0.07589802891016006,
-0.0... |
Unsupervised Learning of Artistic Styles with Archetypal Style Analysis | https://proceedings.neurips.cc/paper_files/paper/2018/hash/09060616068d2b9544dc33f2fbe4ce2d-Abstract.html | [
"Daan Wynen",
"Cordelia Schmid",
"Julien Mairal"
] | null | null | In this paper, we introduce an unsupervised learning approach to automatically dis- cover, summarize, and manipulate artistic styles from large collections of paintings. Our method is based on archetypal analysis, which is an unsupervised learning technique akin to sparse coding with a geometric interpretation. When ap... | [] | null | 41 | 1805.11155 | title_snapshot | [
0.022894226014614105,
-0.02260821871459484,
-0.01583268865942955,
0.023302339017391205,
0.0503804013133049,
0.013093919493258,
0.0027851280756294727,
-0.0277201309800148,
-0.03457197546958923,
-0.07897613197565079,
-0.03998808562755585,
0.0025914879515767097,
-0.07613396644592285,
0.023332... |
Generalisation in humans and deep neural networks | https://proceedings.neurips.cc/paper_files/paper/2018/hash/0937fb5864ed06ffb59ae5f9b5ed67a9-Abstract.html | [
"Robert Geirhos",
"Carlos R. M. Temme",
"Jonas Rauber",
"Heiko H. Schütt",
"Matthias Bethge",
"Felix A. Wichmann"
] | null | null | We compare the robustness of humans and current convolutional deep neural networks (DNNs) on object recognition under twelve different types of image degradations. First, using three well known DNNs (ResNet-152, VGG-19, GoogLeNet) we find the human visual system to be more robust to nearly all of the tested image manip... | [] | null | 42 | 1808.08750 | title_snapshot | [
0.012038089334964752,
0.013403289020061493,
0.010085356421768665,
0.03824548423290253,
0.03193768486380577,
0.02801981009542942,
0.02659590356051922,
0.027493368834257126,
-0.017758401110768318,
-0.05549788475036621,
-0.03020777739584446,
0.008894219994544983,
-0.06738245487213135,
-0.0035... |
Distributed Weight Consolidation: A Brain Segmentation Case Study | https://proceedings.neurips.cc/paper_files/paper/2018/hash/093b60fd0557804c8ba0cbf1453da22f-Abstract.html | [
"Patrick McClure",
"Charles Y Zheng",
"Jakub Kaczmarzyk",
"John Rogers-Lee",
"Satra Ghosh",
"Dylan Nielson",
"Peter A Bandettini",
"Francisco Pereira"
] | null | null | Collecting the large datasets needed to train deep neural networks can be very difficult, particularly for the many applications for which sharing and pooling data is complicated by practical, ethical, or legal concerns. However, it may be the case that derivative datasets or predictive models developed within individu... | [] | null | 43 | 1805.10863 | title_snapshot | [
-0.023010576143860817,
-0.03320342302322388,
-0.015070142224431038,
0.04413409158587456,
0.033109162002801895,
0.04510447382926941,
0.028124799951910973,
-0.0072730910032987595,
-0.04216915741562843,
-0.056001611053943634,
-0.008551215752959251,
-0.01724664494395256,
-0.05918283015489578,
... |
IntroVAE: Introspective Variational Autoencoders for Photographic Image Synthesis | https://proceedings.neurips.cc/paper_files/paper/2018/hash/093f65e080a295f8076b1c5722a46aa2-Abstract.html | [
"Huaibo Huang",
"zhihang li",
"Ran He",
"Zhenan Sun",
"Tieniu Tan"
] | null | null | We present a novel introspective variational autoencoder (IntroVAE) model for synthesizing high-resolution photographic images. IntroVAE is capable of self-evaluating the quality of its generated samples and improving itself accordingly. Its inference and generator models are jointly trained in an introspective way. On... | [] | null | 44 | 1807.06358 | title_snapshot | [
0.02679337002336979,
-0.00426283897832036,
-0.027347231283783913,
0.039422910660505295,
0.035528168082237244,
0.028132479637861252,
0.05210495740175247,
-0.013540706597268581,
-0.04472853243350983,
-0.03907425329089165,
-0.03629929572343826,
-0.011041136458516121,
-0.06204325333237648,
0.0... |
MixLasso: Generalized Mixed Regression via Convex Atomic-Norm Regularization | https://proceedings.neurips.cc/paper_files/paper/2018/hash/09779bb7930c8a0a44360e12b538ae3c-Abstract.html | [
"Ian En-Hsu Yen",
"Wei-Cheng Lee",
"Kai Zhong",
"Sung-En Chang",
"Pradeep K Ravikumar",
"Shou-De Lin"
] | null | null | We consider a generalization of mixed regression where the response is an additive combination of several mixture components. Standard mixed regression is a special case where each response is generated from exactly one component. Typical approaches to the mixture regression problem employ local search methods such as ... | [] | null | 45 | null | null | [
-0.001099507207982242,
-0.0007524931570515037,
-0.008714067749679089,
0.023721808567643166,
0.03389313071966171,
0.04167291894555092,
0.02341023087501526,
-0.01474393717944622,
-0.051430221647024155,
-0.04200693219900131,
-0.022119702771306038,
0.022113822400569916,
-0.06184837594628334,
0... |
A Dual Framework for Low-rank Tensor Completion | https://proceedings.neurips.cc/paper_files/paper/2018/hash/09a5e2a11bea20817477e0b1dfe2cc21-Abstract.html | [
"Madhav Nimishakavi",
"Pratik Kumar Jawanpuria",
"Bamdev Mishra"
] | null | null | One of the popular approaches for low-rank tensor completion is to use the latent trace norm regularization. However, most existing works in this direction learn a sparse combination of tensors. In this work, we fill this gap by proposing a variant of the latent trace norm that helps in learning a non-sparse combinatio... | [] | null | 46 | 1712.01193 | title_snapshot | [
0.0004148151201661676,
-0.02667773887515068,
0.027420302852988243,
0.023404771462082863,
0.0312843956053257,
0.023092476651072502,
0.005047850776463747,
-0.0003389393095858395,
-0.027080751955509186,
-0.07900376617908478,
-0.018519414588809013,
0.019288700073957443,
-0.0272530484944582,
0.... |
Predict Responsibly: Improving Fairness and Accuracy by Learning to Defer | https://proceedings.neurips.cc/paper_files/paper/2018/hash/09d37c08f7b129e96277388757530c72-Abstract.html | [
"David Madras",
"Toni Pitassi",
"Richard Zemel"
] | null | null | In many machine learning applications, there are multiple decision-makers involved, both automated and human. The interaction between these agents often goes unaddressed in algorithmic development. In this work, we explore a simple version of this interaction with a two-stage framework containing an automated model and... | [] | null | 47 | 1711.06664 | title_snapshot | [
-0.005525938235223293,
-0.040845684707164764,
-0.031415991485118866,
0.023923980072140694,
0.04196200519800186,
0.006305405870079994,
0.0297248475253582,
0.021671989932656288,
-0.03261973708868027,
-0.04577362909913063,
0.008850540965795517,
0.03649488836526871,
-0.06670332700014114,
-0.01... |
Enhancing the Accuracy and Fairness of Human Decision Making | https://proceedings.neurips.cc/paper_files/paper/2018/hash/0a113ef6b61820daa5611c870ed8d5ee-Abstract.html | [
"Isabel Valera",
"Adish Singla",
"Manuel Gomez Rodriguez"
] | null | null | Societies often rely on human experts to take a wide variety of decisions affecting their members, from jail-or-release decisions taken by judges and stop-and-frisk decisions taken by police officers to accept-or-reject decisions taken by academics. In this context, each decision is taken by an expert who is typically ... | [] | null | 48 | 1805.10318 | title_snapshot | [
-0.01980295032262802,
0.0037814949173480272,
-0.05382639169692993,
0.053585801273584366,
0.030602166429162025,
0.02813493087887764,
0.015234118327498436,
0.008094865828752518,
-0.021081537008285522,
-0.07908005267381668,
-0.04326196387410164,
0.005878262221813202,
-0.09986674040555954,
-0.... |
Computing Higher Order Derivatives of Matrix and Tensor Expressions | https://proceedings.neurips.cc/paper_files/paper/2018/hash/0a1bf96b7165e962e90cb14648c9462d-Abstract.html | [
"Soeren Laue",
"Matthias Mitterreiter",
"Joachim Giesen"
] | null | null | Optimization is an integral part of most machine learning systems and most numerical optimization schemes rely on the computation of derivatives. Therefore, frameworks for computing derivatives are an active area of machine learning research. Surprisingly, as of yet, no existing framework is capable of computing higher... | [] | null | 49 | null | null | [
-0.042543135583400726,
-0.017077255994081497,
0.017597602680325508,
0.0184563510119915,
0.04116453602910042,
0.027768023312091827,
0.012657038867473602,
-0.014022656716406345,
0.006250872742384672,
-0.030736565589904785,
-0.004761804360896349,
0.04582751914858818,
-0.05422494187951088,
0.0... |
Efficient online algorithms for fast-rate regret bounds under sparsity | https://proceedings.neurips.cc/paper_files/paper/2018/hash/0a348ede8ac3768875037baca5de6e26-Abstract.html | [
"Pierre Gaillard",
"Olivier Wintenberger"
] | null | null | We consider the problem of online convex optimization in two different settings: arbitrary and i.i.d. sequence of convex loss functions. In both settings, we provide efficient algorithms whose cumulative excess risks are controlled with fast-rate sparse bounds. First, the excess risks bounds depend on the sparsity of t... | [] | null | 50 | 1805.09174 | title_snapshot | [
-0.031507931649684906,
-0.007740984205156565,
-0.0055043334141373634,
0.03282732889056206,
0.03199978172779083,
0.032777756452560425,
0.009126300923526287,
0.032538581639528275,
-0.027094466611742973,
-0.021481260657310486,
-0.018826713785529137,
0.009908974170684814,
-0.05625706911087036,
... |
Human-in-the-Loop Interpretability Prior | https://proceedings.neurips.cc/paper_files/paper/2018/hash/0a7d83f084ec258aefd128569dda03d7-Abstract.html | [
"Isaac Lage",
"Andrew Ross",
"Samuel J Gershman",
"Been Kim",
"Finale Doshi-Velez"
] | null | null | We often desire our models to be interpretable as well as accurate. Prior work on optimizing models for interpretability has relied on easy-to-quantify proxies for interpretability, such as sparsity or the number of operations required. In this work, we optimize for interpretability by directly including humans in the ... | [] | null | 51 | 1805.11571 | title_snapshot | [
-0.001432661316357553,
0.0014566112076863647,
-0.020369453355669975,
0.02141018770635128,
0.07453315705060959,
0.007457415573298931,
0.012780995108187199,
-0.004628149792551994,
-0.01703588292002678,
-0.03791357949376106,
-0.04214406758546829,
0.02260502055287361,
-0.060321398079395294,
-0... |
Deep Non-Blind Deconvolution via Generalized Low-Rank Approximation | https://proceedings.neurips.cc/paper_files/paper/2018/hash/0aa1883c6411f7873cb83dacb17b0afc-Abstract.html | [
"Wenqi Ren",
"Jiawei Zhang",
"Lin Ma",
"Jinshan Pan",
"Xiaochun Cao",
"Wangmeng Zuo",
"Wei Liu",
"Ming-Hsuan Yang"
] | null | null | In this paper, we present a deep convolutional neural network to capture the inherent properties of image degradation, which can handle different kernels and saturated pixels in a unified framework. The proposed neural network is motivated by the low-rank property of pseudo-inverse kernels. We first compute a generaliz... | [] | null | 52 | null | null | [
-0.016926394775509834,
-0.018775438889861107,
0.035582683980464935,
0.06027660146355629,
0.03175380453467369,
0.015018834732472897,
0.017471633851528168,
0.0004959931247867644,
-0.025041736662387848,
-0.03743579983711243,
-0.004424160346388817,
0.01924987882375717,
-0.047544095665216446,
0... |
Regret Bounds for Robust Adaptive Control of the Linear Quadratic Regulator | https://proceedings.neurips.cc/paper_files/paper/2018/hash/0ae3f79a30234b6c45a6f7d298ba1310-Abstract.html | [
"Sarah Dean",
"Horia Mania",
"Nikolai Matni",
"Benjamin Recht",
"Stephen Tu"
] | null | null | We consider adaptive control of the Linear Quadratic Regulator (LQR), where an unknown linear system is controlled subject to quadratic costs. Leveraging recent developments in the estimation of linear systems and in robust controller synthesis, we present the first provably polynomial time algorithm that achieves sub-... | [] | null | 53 | 1805.09388 | title_snapshot | [
-0.03725026920437813,
0.0005008013104088604,
-0.007727550342679024,
0.011095613241195679,
0.06783504784107208,
0.028624022379517555,
0.026278894394636154,
-0.0010943451197817922,
-0.014522700570523739,
-0.05264898017048836,
-0.03575723245739937,
0.003722544526681304,
-0.046102214604616165,
... |
Binary Rating Estimation with Graph Side Information | https://proceedings.neurips.cc/paper_files/paper/2018/hash/0b1ec366924b26fc98fa7b71a9c249cf-Abstract.html | [
"Kwangjun Ahn",
"Kangwook Lee",
"Hyunseung Cha",
"Changho Suh"
] | null | null | Rich experimental evidences show that one can better estimate users' unknown ratings with the aid of graph side information such as social graphs. However, the gain is not theoretically quantified. In this work, we study the binary rating estimation problem to understand the fundamental value of graph side information.... | [] | null | 54 | null | null | [
-0.01791340485215187,
-0.007524530868977308,
0.030319364741444588,
0.041872404515743256,
0.008552338927984238,
0.002938780002295971,
0.029910044744610786,
0.015247702598571777,
0.0018010673811659217,
-0.032032888382673264,
0.015637818723917007,
0.015096672810614109,
-0.0676371306180954,
-0... |
A Bayesian Nonparametric View on Count-Min Sketch | https://proceedings.neurips.cc/paper_files/paper/2018/hash/0b9e57c46de934cee33b0e8d1839bfc2-Abstract.html | [
"Diana Cai",
"Michael Mitzenmacher",
"Ryan P. Adams"
] | null | null | The count-min sketch is a time- and memory-efficient randomized data structure that provides a point estimate of the number of times an item has appeared in a data stream. The count-min sketch and related hash-based data structures are ubiquitous in systems that must track frequencies of data such as URLs, IP addresses... | [] | null | 55 | null | null | [
-0.0003605168021749705,
-0.012168069370090961,
-0.015662623569369316,
0.04535725340247154,
0.016734719276428223,
0.03558064252138138,
0.03020421974360943,
0.009049857966601849,
-0.02800453081727028,
-0.06461799889802933,
-0.018609261140227318,
-0.017271999269723892,
-0.0458260215818882,
-0... |
Provable Variational Inference for Constrained Log-Submodular Models | https://proceedings.neurips.cc/paper_files/paper/2018/hash/0c0a7566915f4f24853fc4192689aa7e-Abstract.html | [
"Josip Djolonga",
"Stefanie Jegelka",
"Andreas Krause"
] | null | null | Submodular maximization problems appear in several areas of machine learning and data science, as many useful modelling concepts such as diversity and coverage satisfy this natural diminishing returns property. Because the data defining these functions, as well as the decisions made with the computed solutions, are sub... | [] | null | 56 | null | null | [
-0.010298257693648338,
-0.006913842633366585,
-0.012948212213814259,
0.048097264021635056,
0.06933145970106125,
0.03672662749886513,
0.028591688722372055,
-0.014626026153564453,
-0.024211861193180084,
-0.01756233163177967,
-0.012680492363870144,
0.008344849571585655,
-0.07878807932138443,
... |
Middle-Out Decoding | https://proceedings.neurips.cc/paper_files/paper/2018/hash/0c215f194276000be6a6df6528067151-Abstract.html | [
"Shikib Mehri",
"Leonid Sigal"
] | null | null | Despite being virtually ubiquitous, sequence-to-sequence models are challenged by their lack of diversity and inability to be externally controlled. In this paper, we speculate that a fundamental shortcoming of sequence generation models is that the decoding is done strictly from left-to-right, meaning that outputs val... | [] | null | 57 | 1810.11735 | title_snapshot | [
0.007835740223526955,
-0.03177817538380623,
-0.008744236081838608,
0.02248748019337654,
0.024066904559731483,
0.01809859648346901,
0.02011081390082836,
0.026387885212898254,
-0.027561863884329796,
-0.013813149183988571,
-0.023045875132083893,
0.008783831261098385,
-0.053936608135700226,
-0... |
Maximum-Entropy Fine Grained Classification | https://proceedings.neurips.cc/paper_files/paper/2018/hash/0c74b7f78409a4022a2c4c5a5ca3ee19-Abstract.html | [
"Abhimanyu Dubey",
"Otkrist Gupta",
"Ramesh Raskar",
"Nikhil Naik"
] | null | null | Fine-Grained Visual Classification (FGVC) is an important computer vision problem that involves small diversity within the different classes, and often requires expert annotators to collect data. Utilizing this notion of small visual diversity, we revisit Maximum-Entropy learning in the context of fine-grained classifi... | [] | null | 58 | 1809.05934 | title_snapshot | [
0.006668770220130682,
-0.0218605138361454,
0.004127140622586012,
0.043807417154312134,
0.026615092530846596,
0.027244435623288155,
0.009216020815074444,
0.012001881375908852,
-0.03198631852865219,
-0.05933032184839249,
-0.030918775126338005,
0.009450903162360191,
-0.06803928315639496,
0.01... |
Deep State Space Models for Unconditional Word Generation | https://proceedings.neurips.cc/paper_files/paper/2018/hash/0cd60efb5578cd967c3c23894f305800-Abstract.html | [
"Florian Schmidt",
"Thomas Hofmann"
] | null | null | Autoregressive feedback is considered a necessity for successful unconditional text generation using stochastic sequence models. However, such feedback is known to introduce systematic biases into the training process and it obscures a principle of generation: committing to global information and forgetting local nuanc... | [] | null | 59 | 1806.04550 | title_snapshot | [
-0.009147357195615768,
-0.023113546893000603,
-0.02344118431210518,
0.06896878778934479,
0.04967283457517624,
0.053965386003255844,
0.03601715713739395,
0.027946636080741882,
0.0008852996397763491,
-0.03401540219783783,
-0.010492820292711258,
0.010276423767209053,
-0.07494072616100311,
-0.... |
Total stochastic gradient algorithms and applications in reinforcement learning | https://proceedings.neurips.cc/paper_files/paper/2018/hash/0d59701b3474225fca5563e015965886-Abstract.html | [
"Paavo Parmas"
] | null | null | Backpropagation and the chain rule of derivatives have been prominent; however, the total derivative rule has not enjoyed the same amount of attention. In this work we show how the total derivative rule leads to an intuitive visual framework for creating gradient estimators on graphical models. In particular, previous ... | [] | null | 60 | 1902.01722 | title_snapshot | [
-0.030363503843545914,
-0.011432240717113018,
-0.005329920910298824,
0.03659207001328468,
0.02479623258113861,
0.03141571581363678,
0.026302838698029518,
0.03235705941915512,
-0.02051478624343872,
-0.03533567115664482,
0.00242919335141778,
0.0078820725902915,
-0.06809744983911514,
0.001418... |
Neural Arithmetic Logic Units | https://proceedings.neurips.cc/paper_files/paper/2018/hash/0e64a7b00c83e3d22ce6b3acf2c582b6-Abstract.html | [
"Andrew Trask",
"Felix Hill",
"Scott E Reed",
"Jack Rae",
"Chris Dyer",
"Phil Blunsom"
] | null | null | Neural networks can learn to represent and manipulate numerical information, but they seldom generalize well outside of the range of numerical values encountered during training. To encourage more systematic numerical extrapolation, we propose an architecture that represents numerical quantities as linear activations w... | [] | null | 61 | 1808.00508 | title_snapshot | [
-0.009716837666928768,
0.035477783530950546,
-0.00907812174409628,
0.025446781888604164,
0.04235425591468811,
0.05415226146578789,
0.020081115886569023,
0.004903677385300398,
-0.07691484689712524,
0.003987116273492575,
0.0026885755360126495,
0.0004283655434846878,
-0.050374019891023636,
0.... |
Implicit Bias of Gradient Descent on Linear Convolutional Networks | https://proceedings.neurips.cc/paper_files/paper/2018/hash/0e98aeeb54acf612b9eb4e48a269814c-Abstract.html | [
"Suriya Gunasekar",
"Jason Lee",
"Daniel Soudry",
"Nati Srebro"
] | null | null | We show that gradient descent on full-width linear convolutional networks of depth $L$ converges to a linear predictor related to the $\ell_{2/L}$ bridge penalty in the frequency domain. This is in contrast to linearly fully connected networks, where gradient descent converges to the hard margin linear SVM solution, re... | [] | null | 62 | 1806.00468 | title_snapshot | [
-0.0006305380375124514,
-0.01902114227414131,
0.012827838771045208,
0.01708989031612873,
0.025082815438508987,
0.014448913745582104,
0.03716845065355301,
0.03834426775574684,
-0.022262219339609146,
-0.029974626377224922,
0.014251336455345154,
0.005592007655650377,
-0.08366542309522629,
0.0... |
On Binary Classification in Extreme Regions | https://proceedings.neurips.cc/paper_files/paper/2018/hash/0ebcc77dc72360d0eb8e9504c78d38bd-Abstract.html | [
"Hamid JALALZAI",
"Stephan Clémençon",
"Anne Sabourin"
] | null | null | In pattern recognition, a random label Y is to be predicted based upon observing a random vector X valued in $\mathbb{R}^d$ with d>1 by means of a classification rule with minimum probability of error. In a wide variety of applications, ranging from finance/insurance to environmental sciences through teletraffic data a... | [] | null | 63 | null | null | [
-0.024134185165166855,
-0.021920762956142426,
-0.004067635629326105,
0.0436619371175766,
0.0368344709277153,
0.02725379168987274,
0.007894052192568779,
0.008569139987230301,
-0.02883417159318924,
-0.025320515036582947,
-0.04349958151578903,
-0.0023672119714319706,
-0.07127974927425385,
0.0... |
Adaptive Skip Intervals: Temporal Abstraction for Recurrent Dynamical Models | https://proceedings.neurips.cc/paper_files/paper/2018/hash/0f0ee3310223fe38a989b2c818709393-Abstract.html | [
"Alexander Neitz",
"Giambattista Parascandolo",
"Stefan Bauer",
"Bernhard Schölkopf"
] | null | null | We introduce a method which enables a recurrent dynamics model to be temporally abstract. Our approach, which we call Adaptive Skip Intervals (ASI), is based on the observation that in many sequential prediction tasks, the exact time at which events occur is irrelevant to the underlying objective. Moreover, in many sit... | [] | null | 64 | 1808.04768 | title_snapshot | [
-0.05379908159375191,
-0.01826685294508934,
-0.022222287952899933,
0.025398027151823044,
0.03312813118100166,
0.004140973556786776,
0.05446162819862366,
0.035488929599523544,
-0.05259004980325699,
-0.025295400992035866,
-0.004537129309028387,
0.012012984603643417,
-0.07371198385953903,
-0.... |
Learning to Specialize with Knowledge Distillation for Visual Question Answering | https://proceedings.neurips.cc/paper_files/paper/2018/hash/0f2818101a7ac4b96ceeba38de4b934c-Abstract.html | [
"Jonghwan Mun",
"Kimin Lee",
"Jinwoo Shin",
"Bohyung Han"
] | null | null | Visual Question Answering (VQA) is a notoriously challenging problem because it involves various heterogeneous tasks defined by questions within a unified framework. Learning specialized models for individual types of tasks is intuitively attracting but surprisingly difficult; it is not straightforward to outperform na... | [] | null | 65 | null | null | [
0.02341439016163349,
-0.0065634530037641525,
0.008679633028805256,
0.06798817962408066,
0.04683360829949379,
0.015368398278951645,
0.015104711055755615,
-0.010873808525502682,
-0.03217725083231926,
0.0005804173415526748,
-0.050599269568920135,
0.04097602143883705,
-0.06837256997823715,
-0.... |
A Model for Learned Bloom Filters and Optimizing by Sandwiching | https://proceedings.neurips.cc/paper_files/paper/2018/hash/0f49c89d1e7298bb9930789c8ed59d48-Abstract.html | [
"Michael Mitzenmacher"
] | null | null | Recent work has suggested enhancing Bloom filters by using a pre-filter, based on applying machine learning to determine a function that models the data set the Bloom filter is meant to represent. Here we model such learned Bloom filters, with the following outcomes: (1) we clarify what guarantees can and cannot be ass... | [] | null | 66 | 1901.00902 | title_snapshot | [
0.004574653692543507,
-0.04159805551171303,
0.006804667878895998,
0.039938196539878845,
0.07107771933078766,
-0.006872242316603661,
0.004126198589801788,
-0.01809440180659294,
-0.038401782512664795,
-0.03613239526748657,
0.008323556743562222,
-0.03204801678657532,
-0.040423933416604996,
0.... |
Random Feature Stein Discrepancies | https://proceedings.neurips.cc/paper_files/paper/2018/hash/0f840be9b8db4d3fbd5ba2ce59211f55-Abstract.html | [
"Jonathan Huggins",
"Lester Mackey"
] | null | null | Computable Stein discrepancies have been deployed for a variety of applications, ranging from sampler selection in posterior inference to approximate Bayesian inference to goodness-of-fit testing. Existing convergence-determining Stein discrepancies admit strong theoretical guarantees but suffer from a computational co... | [] | null | 67 | 1806.07788 | title_snapshot | [
-0.013416122645139694,
0.0058082896284759045,
-0.013267328962683678,
0.06508579105138779,
0.04741401970386505,
0.04496998339891434,
0.027082394808530807,
-0.024529989808797836,
-0.015071101486682892,
-0.050386346876621246,
0.003153072902932763,
-0.010962523519992828,
-0.06017926707863808,
... |
Nonparametric Bayesian Lomax delegate racing for survival analysis with competing risks | https://proceedings.neurips.cc/paper_files/paper/2018/hash/0fe473396242072e84af286632d3f0ff-Abstract.html | [
"Quan Zhang",
"Mingyuan Zhou"
] | null | null | We propose Lomax delegate racing (LDR) to explicitly model the mechanism of survival under competing risks and to interpret how the covariates accelerate or decelerate the time to event. LDR explains non-monotonic covariate effects by racing a potentially infinite number of sub-risks, and consequently relaxes the ubiqu... | [] | null | 68 | 1810.08564 | title_snapshot | [
-0.014014901593327522,
-0.03433288633823395,
-0.041930925101041794,
0.019734319299459457,
0.02617761865258217,
0.04189907759428024,
0.0229695625603199,
-0.03028678148984909,
-0.02689204178750515,
-0.03905623033642769,
0.03171498328447342,
0.0054021961987018585,
-0.044937629252672195,
0.012... |
Hessian-based Analysis of Large Batch Training and Robustness to Adversaries | https://proceedings.neurips.cc/paper_files/paper/2018/hash/102f0bb6efb3a6128a3c750dd16729be-Abstract.html | [
"Zhewei Yao",
"Amir Gholami",
"Qi Lei",
"Kurt Keutzer",
"Michael W. Mahoney"
] | null | null | Large batch size training of Neural Networks has been shown to incur accuracy loss when trained with the current methods. The exact underlying reasons for this are still not completely understood. Here, we study large batch size training through the lens of the Hessian operator and robust optimization. In particular, w... | [] | null | 69 | 1802.08241 | title_snapshot | [
-0.05572469159960747,
-0.03393245488405228,
-0.004594456870108843,
0.03567267954349518,
0.01551692932844162,
0.026768842712044716,
0.0214239452034235,
0.005979107692837715,
-0.04089376702904701,
-0.061107337474823,
0.008343051187694073,
-0.003947553224861622,
-0.057584136724472046,
-0.0030... |
Adaptive Online Learning in Dynamic Environments | https://proceedings.neurips.cc/paper_files/paper/2018/hash/10a5ab2db37feedfdeaab192ead4ac0e-Abstract.html | [
"Lijun Zhang",
"Shiyin Lu",
"Zhi-Hua Zhou"
] | null | null | In this paper, we study online convex optimization in dynamic environments, and aim to bound the dynamic regret with respect to any sequence of comparators. Existing work have shown that online gradient descent enjoys an $O(\sqrt{T}(1+P_T))$ dynamic regret, where $T$ is the number of iterations and $P_T$ is the path-le... | [] | null | 70 | 1810.10815 | title_snapshot | [
-0.04547076299786568,
-0.005772857461124659,
0.01457914337515831,
0.03357834368944168,
0.042623668909072876,
0.048767559230327606,
0.018021738156676292,
0.02029886096715927,
-0.043457649648189545,
-0.03959188610315323,
-0.021085405722260475,
0.0117535050958395,
-0.037665955722332,
-0.02228... |
Learning in Games with Lossy Feedback | https://proceedings.neurips.cc/paper_files/paper/2018/hash/10c66082c124f8afe3df4886f5e516e0-Abstract.html | [
"Zhengyuan Zhou",
"Panayotis Mertikopoulos",
"Susan Athey",
"Nicholas Bambos",
"Peter W. Glynn",
"Yinyu Ye"
] | null | null | We consider a game-theoretical multi-agent learning problem where the feedback information can be lost during the learning process and rewards are given by a broad class of games known as variationally stable games. We propose a simple variant of the classical online gradient descent algorithm, called reweighted online... | [] | null | 71 | null | null | [
-0.05023057386279106,
-0.015127263963222504,
0.022890573367476463,
0.03966420143842697,
0.03147554397583008,
0.02407771348953247,
0.005424685310572386,
0.028051480650901794,
-0.036897700279951096,
-0.03245622664690018,
-0.016847413033246994,
0.030606118962168694,
-0.06998732686042786,
-0.0... |
Statistical Optimality of Stochastic Gradient Descent on Hard Learning Problems through Multiple Passes | https://proceedings.neurips.cc/paper_files/paper/2018/hash/10ff0b5e85e5b85cc3095d431d8c08b4-Abstract.html | [
"Loucas Pillaud-Vivien",
"Alessandro Rudi",
"Francis Bach"
] | null | null | We consider stochastic gradient descent (SGD) for least-squares regression with potentially several passes over the data. While several passes have been widely reported to perform practically better in terms of predictive performance on unseen data, the existing theoretical analysis of SGD suggests that a single pass i... | [] | null | 72 | 1805.10074 | title_snapshot | [
-0.04777500778436661,
-0.008573046885430813,
0.029054708778858185,
0.036110468208789825,
0.04884183779358864,
0.0377272292971611,
0.037723690271377563,
-0.014665118418633938,
-0.02768583782017231,
-0.02705681324005127,
-0.032432593405246735,
0.004981282167136669,
-0.0391668975353241,
0.012... |
Multimodal Generative Models for Scalable Weakly-Supervised Learning | https://proceedings.neurips.cc/paper_files/paper/2018/hash/1102a326d5f7c9e04fc3c89d0ede88c9-Abstract.html | [
"Mike Wu",
"Noah Goodman"
] | null | null | Multiple modalities often co-occur when describing natural phenomena. Learning a joint representation of these modalities should yield deeper and more useful representations.Previous generative approaches to multi-modal input either do not learn a joint distribution or require additional computation to handle missing d... | [] | null | 73 | 1802.05335 | title_snapshot | [
0.01822371408343315,
-0.002735992195084691,
-0.0042364345863461494,
0.06683369725942612,
0.02067681960761547,
0.05113891139626503,
0.047199755907058716,
0.03131774440407753,
-0.034304507076740265,
-0.02513130009174347,
-0.024426061660051346,
0.02529684081673622,
-0.07538792490959167,
0.022... |
Multi-Class Learning: From Theory to Algorithm | https://proceedings.neurips.cc/paper_files/paper/2018/hash/1141938ba2c2b13f5505d7c424ebae5f-Abstract.html | [
"Jian Li",
"Yong Liu",
"Rong Yin",
"Hua Zhang",
"Lizhong Ding",
"Weiping Wang"
] | null | null | In this paper, we study the generalization performance of multi-class classification and obtain a shaper data-dependent generalization error bound with fast convergence rate, substantially improving the state-of-art bounds in the existing data-dependent generalization analysis. The theoretical analysis motivates us to ... | [] | null | 74 | null | null | [
-0.018705209717154503,
-0.02292662113904953,
0.021602459251880646,
0.04464630410075188,
0.03587515652179718,
0.06510971486568451,
-0.0026417234912514687,
-0.031038997694849968,
-0.026784630492329597,
-0.03097539022564888,
-0.014358232729136944,
0.03146488219499588,
-0.07339172065258026,
0.... |
Learning and Inference in Hilbert Space with Quantum Graphical Models | https://proceedings.neurips.cc/paper_files/paper/2018/hash/11704817e347269b7254e744b5e22dac-Abstract.html | [
"Siddarth Srinivasan",
"Carlton Downey",
"Byron Boots"
] | null | null | Quantum Graphical Models (QGMs) generalize classical graphical models by adopting the formalism for reasoning about uncertainty from quantum mechanics. Unlike classical graphical models, QGMs represent uncertainty with density matrices in complex Hilbert spaces. Hilbert space embeddings (HSEs) also generalize Bayesian ... | [] | null | 75 | 1810.12369 | title_snapshot | [
-0.02884078025817871,
0.024303114041686058,
-0.021006189286708832,
0.04324889928102493,
0.044273484498262405,
-0.016218477860093117,
0.04017806425690651,
-0.0028419424779713154,
0.007552192080765963,
-0.0559241846203804,
-0.02175256423652172,
0.0197959803044796,
-0.0877305120229721,
0.0143... |
Bayesian Structure Learning by Recursive Bootstrap | https://proceedings.neurips.cc/paper_files/paper/2018/hash/11e2ad6bf99300cd3808bb105b55d4b8-Abstract.html | [
"Raanan Y. Rohekar",
"Yaniv Gurwicz",
"Shami Nisimov",
"Guy Koren",
"Gal Novik"
] | null | null | We address the problem of Bayesian structure learning for domains with hundreds of variables by employing non-parametric bootstrap, recursively. We propose a method that covers both model averaging and model selection in the same framework. The proposed method deals with the main weakness of constraint-based learning--... | [] | null | 76 | 1809.04828 | title_snapshot | [
-0.01751095987856388,
0.009355816058814526,
-0.014895506203174591,
0.041600070893764496,
0.030743226408958435,
0.03676096349954605,
0.052119962871074677,
-0.009323390200734138,
-0.014023992232978344,
-0.038301412016153336,
0.002670754212886095,
0.02175949513912201,
-0.08571654558181763,
-0... |
Efficient Convex Completion of Coupled Tensors using Coupled Nuclear Norms | https://proceedings.neurips.cc/paper_files/paper/2018/hash/12092a75caa75e4644fd2869f0b6c45a-Abstract.html | [
"Kishan Wimalawarne",
"Hiroshi Mamitsuka"
] | null | null | Coupled norms have emerged as a convex method to solve coupled tensor completion. A limitation with coupled norms is that they only induce low-rankness using the multilinear rank of coupled tensors. In this paper, we introduce a new set of coupled norms known as coupled nuclear norms by constraining the CP rank of coup... | [] | null | 77 | null | null | [
-0.0002871365868486464,
-0.01663956791162491,
0.02397466078400612,
0.024303510785102844,
0.009847559034824371,
0.0001138529769377783,
0.0009550440590828657,
-0.009323691949248314,
-0.04331936687231064,
-0.06324096769094467,
-0.048755403608083725,
0.020225942134857178,
-0.017752181738615036,
... |
Stein Variational Gradient Descent as Moment Matching | https://proceedings.neurips.cc/paper_files/paper/2018/hash/125b93c9b50703fe9dac43ec231f5f83-Abstract.html | [
"Qiang Liu",
"Dilin Wang"
] | null | null | Stein variational gradient descent (SVGD) is a non-parametric inference algorithm that evolves a set of particles to fit a given distribution of interest. We analyze the non-asymptotic properties of SVGD, showing that there exists a set of functions, which we call the Stein matching set, whose expectations are exactly ... | [] | null | 78 | 1810.11693 | title_snapshot | [
-0.028492683544754982,
0.004368826746940613,
0.027109526097774506,
0.03145836666226387,
0.02052297629415989,
0.04936430603265762,
0.016493666917085648,
-0.0030860307160764933,
-0.02578124590218067,
-0.038530442863702774,
-0.008960963226854801,
-0.003944692201912403,
-0.0605282261967659,
-0... |
Data-Driven Clustering via Parameterized Lloyd's Families | https://proceedings.neurips.cc/paper_files/paper/2018/hash/128ac9c427302b7a64314fc4593430b2-Abstract.html | [
"Maria-Florina F Balcan",
"Travis Dick",
"Colin White"
] | null | null | Algorithms for clustering points in metric spaces is a long-studied area of research. Clustering has seen a multitude of work both theoretically, in understanding the approximation guarantees possible for many objective functions such as k-median and k-means clustering, and experimentally, in finding the fastest algori... | [] | null | 79 | 1809.06987 | title_snapshot | [
-0.01088019646704197,
-0.003437773324549198,
0.013144243508577347,
0.027994602918624878,
0.0505945198237896,
0.05588391050696373,
0.014153245836496353,
-0.003245583502575755,
0.0023954648058861494,
-0.03475804254412651,
-0.02189088799059391,
-0.008426767773926258,
-0.05919995158910751,
0.0... |
Semi-Supervised Learning with Declaratively Specified Entropy Constraints | https://proceedings.neurips.cc/paper_files/paper/2018/hash/12b1e42dc0746f22cf361267de07073f-Abstract.html | [
"Haitian Sun",
"William W. Cohen",
"Lidong Bing"
] | null | null | We propose a technique for declaratively specifying strategies for semi-supervised learning (SSL). SSL methods based on different assumptions perform differently on different tasks, which leads to difficulties applying them in practice. In this paper, we propose to use entropy to unify many types of constraints. Our me... | [] | null | 80 | 1804.09238 | title_snapshot | [
0.008689398877322674,
-0.04506230354309082,
-0.029604006558656693,
0.03207970783114433,
0.05540989339351654,
0.015537810511887074,
0.019881820306181908,
-0.030801542103290558,
-0.026499493047595024,
-0.01698073372244835,
-0.05908208340406418,
0.03750770166516304,
-0.05854359641671181,
0.01... |
Bayesian Inference of Temporal Task Specifications from Demonstrations | https://proceedings.neurips.cc/paper_files/paper/2018/hash/13168e6a2e6c84b4b7de9390c0ef5ec5-Abstract.html | [
"Ankit Shah",
"Pritish Kamath",
"Julie A Shah",
"Shen Li"
] | null | null | When observing task demonstrations, human apprentices are able to identify whether a given task is executed correctly long before they gain expertise in actually performing that task. Prior research into learning from demonstrations (LfD) has failed to capture this notion of the acceptability of an execution; meanwhile... | [] | null | 81 | null | null | [
0.014157354831695557,
0.023601563647389412,
-0.04624766483902931,
0.033140748739242554,
0.0469837412238121,
-0.003964163362979889,
0.030256206169724464,
0.011640708893537521,
-0.016661521047353745,
-0.001308348379097879,
-0.02805010788142681,
0.06394588947296143,
-0.042490649968385696,
-0.... |
Stochastic Nested Variance Reduction for Nonconvex Optimization | https://proceedings.neurips.cc/paper_files/paper/2018/hash/136f951362dab62e64eb8e841183c2a9-Abstract.html | [
"Dongruo Zhou",
"Pan Xu",
"Quanquan Gu"
] | null | null | We study finite-sum nonconvex optimization problems, where the objective function is an average of $n$ nonconvex functions. We propose a new stochastic gradient descent algorithm based on nested variance reduction. Compared with conventional stochastic variance reduced gradient (SVRG) algorithm that uses two reference ... | [] | null | 82 | 1806.07811 | title_snapshot | [
-0.012697170488536358,
-0.01072128489613533,
0.013899551704525948,
0.02751961164176464,
0.01907072588801384,
0.05747276172041893,
0.02788885496556759,
0.0007228310569189489,
-0.007551901042461395,
-0.052252721041440964,
0.00853051245212555,
-0.032617196440696716,
-0.04984928295016289,
0.00... |
On Markov Chain Gradient Descent | https://proceedings.neurips.cc/paper_files/paper/2018/hash/1371bccec2447b5aa6d96d2a540fb401-Abstract.html | [
"Tao Sun",
"Yuejiao Sun",
"Wotao Yin"
] | null | null | Stochastic gradient methods are the workhorse (algorithms) of large-scale optimization problems in machine learning, signal processing, and other computational sciences and engineering. This paper studies Markov chain gradient descent, a variant of stochastic gradient descent where the random samples are taken on the t... | [] | null | 83 | 1809.04216 | title_snapshot | [
-0.03347639739513397,
-0.008055471815168858,
-0.01066338736563921,
0.03622714430093765,
0.052613258361816406,
0.024708352982997894,
0.036579616367816925,
0.018243832513689995,
-0.011361081153154373,
-0.0507410503923893,
0.0010786015773192048,
-0.004998407792299986,
-0.05218356475234032,
-0... |
The Limit Points of (Optimistic) Gradient Descent in Min-Max Optimization | https://proceedings.neurips.cc/paper_files/paper/2018/hash/139c3c1b7ca46a9d4fd6d163d98af635-Abstract.html | [
"Constantinos Daskalakis",
"Ioannis Panageas"
] | null | null | Motivated by applications in Optimization, Game Theory, and the training of Generative Adversarial Networks, the convergence properties of first order methods in min-max problems have received extensive study. It has been recognized that they may cycle, and there is no good understanding of their limit points when they... | [] | null | 84 | 1807.03907 | title_snapshot | [
-0.05661608278751373,
-0.02227034978568554,
0.010072815231978893,
0.03607961907982826,
0.019598089158535004,
0.03085682913661003,
0.02355976216495037,
0.013836698606610298,
-0.03357578068971634,
-0.043164853006601334,
-0.006721530109643936,
-0.004164078272879124,
-0.07529420405626297,
0.00... |
Empirical Risk Minimization in Non-interactive Local Differential Privacy Revisited | https://proceedings.neurips.cc/paper_files/paper/2018/hash/13f320e7b5ead1024ac95c3b208610db-Abstract.html | [
"Di Wang",
"Marco Gaboardi",
"Jinhui Xu"
] | null | null | In this paper, we revisit the Empirical Risk Minimization problem in the non-interactive local model of differential privacy. In the case of constant or low dimensions ($p\ll n$), we first show that if the loss function is $(\infty, T)$-smooth, we can avoid a dependence of the sample complexity, to achieve error $\alph... | [] | null | 85 | 1802.04085 | title_judge | [
-0.03874940425157547,
-0.0018253745511174202,
0.017064008861780167,
0.05325894430279732,
0.04673408716917038,
0.041509632021188736,
0.03760360926389694,
-0.031468797475099564,
-0.01893031783401966,
-0.03740508109331131,
0.00827158335596323,
-0.023165274411439896,
-0.06455470621585846,
-0.0... |
Greedy Hash: Towards Fast Optimization for Accurate Hash Coding in CNN | https://proceedings.neurips.cc/paper_files/paper/2018/hash/13f3cf8c531952d72e5847c4183e6910-Abstract.html | [
"Shupeng Su",
"Chao Zhang",
"Kai Han",
"Yonghong Tian"
] | null | null | To convert the input into binary code, hashing algorithm has been widely used for approximate nearest neighbor search on large-scale image sets due to its computation and storage efficiency. Deep hashing further improves the retrieval quality by combining the hash coding with deep neural network. However, a major diffi... | [] | null | 86 | null | null | [
-0.028692569583654404,
-0.01687045954167843,
-0.025814218446612358,
0.05763430520892143,
0.03854582831263542,
0.0246279239654541,
0.008382698521018028,
0.005119693465530872,
-0.02268216758966446,
-0.044165562838315964,
-0.02474161423742771,
-0.029872952029109,
-0.044512148946523666,
-0.010... |
Which Neural Net Architectures Give Rise to Exploding and Vanishing Gradients? | https://proceedings.neurips.cc/paper_files/paper/2018/hash/13f9896df61279c928f19721878fac41-Abstract.html | [
"Boris Hanin"
] | null | null | We give a rigorous analysis of the statistical behavior of gradients in a randomly initialized fully connected network N with ReLU activations. Our results show that the empirical variance of the squares of the entries in the input-output Jacobian of N is exponential in a simple architecture-dependent constant beta, gi... | [] | null | 87 | 1801.03744 | title_snapshot | [
-0.007377306465059519,
-0.005106986965984106,
0.004684892483055592,
0.04527384787797928,
0.032078348100185394,
0.037895701825618744,
0.024706948548555374,
0.025134438648819923,
-0.053625497967004776,
-0.01842465251684189,
0.0029565985314548016,
-0.023747894912958145,
-0.07000122964382172,
... |
Learning a High Fidelity Pose Invariant Model for High-resolution Face Frontalization | https://proceedings.neurips.cc/paper_files/paper/2018/hash/1415db70fe9ddb119e23e9b2808cde38-Abstract.html | [
"Jie Cao",
"Yibo Hu",
"Hongwen Zhang",
"Ran He",
"Zhenan Sun"
] | null | null | Face frontalization refers to the process of synthesizing the frontal view of a face from a given profile. Due to self-occlusion and appearance distortion in the wild, it is extremely challenging to recover faithful results and preserve texture details in a high-resolution. This paper proposes a High Fidelity Pose Inva... | [] | null | 88 | 1806.08472 | title_snapshot | [
-0.03801450878381729,
0.003019333351403475,
-0.010201387107372284,
0.01061844639480114,
0.026653647422790527,
0.0406217947602272,
0.019090991467237473,
0.02370859682559967,
-0.027038414031267166,
-0.05508119985461235,
-0.013640513643622398,
-0.01961406320333481,
-0.04288019239902496,
-0.00... |
Provably Correct Automatic Sub-Differentiation for Qualified Programs | https://proceedings.neurips.cc/paper_files/paper/2018/hash/142c65e00f4f7cf2e6c4c996e34005df-Abstract.html | [
"Sham M. Kakade",
"Jason Lee"
] | null | null | The \emph{Cheap Gradient Principle}~\citep{Griewank:2008:EDP:1455489} --- the computational cost of computing a $d$-dimensional vector of partial derivatives of a scalar function is nearly the same (often within a factor of $5$) as that of simply computing the scalar function itself --- is of central importance in opti... | [] | null | 89 | 1809.08530 | title_judge | [
-0.04668543115258217,
-0.02693857066333294,
0.0173332542181015,
0.023362137377262115,
0.050693415105342865,
0.03373372554779053,
0.02830306440591812,
-0.004494634922593832,
0.0035975545179098845,
-0.03873264044523239,
0.0032644509337842464,
0.008420332334935665,
-0.05265447869896889,
0.016... |
CatBoost: unbiased boosting with categorical features | https://proceedings.neurips.cc/paper_files/paper/2018/hash/14491b756b3a51daac41c24863285549-Abstract.html | [
"Liudmila Prokhorenkova",
"Gleb Gusev",
"Aleksandr Vorobev",
"Anna Veronika Dorogush",
"Andrey Gulin"
] | null | null | This paper presents the key algorithmic techniques behind CatBoost, a new gradient boosting toolkit. Their combination leads to CatBoost outperforming other publicly available boosting implementations in terms of quality on a variety of datasets. Two critical algorithmic advances introduced in CatBoost are the implemen... | [] | null | 90 | 1706.09516 | title_snapshot | [
-0.017941482365131378,
-0.06084081158041954,
0.0012879340210929513,
0.015946675091981888,
0.04002365469932556,
0.008690709248185158,
0.01718151941895485,
0.008908767253160477,
-0.009264660999178886,
-0.033806368708610535,
-0.04343312978744507,
0.0020144833251833916,
-0.06723231077194214,
-... |
Constrained Generation of Semantically Valid Graphs via Regularizing Variational Autoencoders | https://proceedings.neurips.cc/paper_files/paper/2018/hash/1458e7509aa5f47ecfb92536e7dd1dc7-Abstract.html | [
"Tengfei Ma",
"Jie Chen",
"Cao Xiao"
] | null | null | Deep generative models have achieved remarkable success in various data domains, including images, time series, and natural languages. There remain, however, substantial challenges for combinatorial structures, including graphs. One of the key challenges lies in the difficulty of ensuring semantic validity in context. ... | [] | null | 91 | 1809.02630 | title_snapshot | [
0.006683288607746363,
-0.024022065103054047,
-0.025584178045392036,
0.06394420564174652,
0.049051422625780106,
0.025304563343524933,
0.035264525562524796,
0.009612544439733028,
0.0008748773252591491,
-0.05163605511188507,
-0.0026487235445529222,
0.010282515548169613,
-0.08258912712335587,
... |
Deep, complex, invertible networks for inversion of transmission effects in multimode optical fibres | https://proceedings.neurips.cc/paper_files/paper/2018/hash/148510031349642de5ca0c544f31b2ef-Abstract.html | [
"Oisín Moran",
"Piergiorgio Caramazza",
"Daniele Faccio",
"Roderick Murray-Smith"
] | null | null | We use complex-weighted, deep networks to invert the effects of multimode optical fibre distortion of a coherent input image. We generated experimental data based on collections of optical fibre responses to greyscale input images generated with coherent light, by measuring only image amplitude (not amplitude and phase... | [] | null | 92 | null | null | [
0.00395223218947649,
-0.027518142014741898,
0.03527110815048218,
0.026581499725580215,
0.041010044515132904,
0.02184237167239189,
0.01847236417233944,
0.009543917141854763,
-0.022663643583655357,
-0.044631801545619965,
0.0013510987628251314,
-0.00878939125686884,
-0.0461772158741951,
0.023... |
Scalable Hyperparameter Transfer Learning | https://proceedings.neurips.cc/paper_files/paper/2018/hash/14c879f3f5d8ed93a09f6090d77c2cc3-Abstract.html | [
"Valerio Perrone",
"Rodolphe Jenatton",
"Matthias W Seeger",
"Cedric Archambeau"
] | null | null | Bayesian optimization (BO) is a model-based approach for gradient-free black-box function optimization, such as hyperparameter optimization. Typically, BO relies on conventional Gaussian process (GP) regression, whose algorithmic complexity is cubic in the number of evaluations. As a result, GP-based BO cannot leverage... | [] | null | 93 | null | null | [
-0.010569135658442974,
-0.004297229461371899,
-0.0031034708954393864,
0.01187556330114603,
0.030321529135107994,
0.04661141335964203,
0.02459435909986496,
-0.02440415322780609,
0.015863453969359398,
-0.03205738961696625,
-0.003727253759279847,
0.03814702108502388,
-0.043316349387168884,
0.... |
Wasserstein Distributionally Robust Kalman Filtering | https://proceedings.neurips.cc/paper_files/paper/2018/hash/15212f24321aa2c3dc8e9acf820f3c15-Abstract.html | [
"Soroosh Shafieezadeh-Abadeh",
"Viet Anh Nguyen",
"Daniel Huhn",
"Peyman Mohajerin Esfahani"
] | null | null | We study a distributionally robust mean square error estimation problem over a nonconvex Wasserstein ambiguity set containing only normal distributions. We show that the optimal estimator and the least favorable distribution form a Nash equilibrium. Despite the non-convex nature of the ambiguity set, we prove that the ... | [] | null | 94 | 1809.08830 | title_snapshot | [
-0.024261778220534325,
0.02502787485718727,
0.021146319806575775,
0.02861754782497883,
0.03538646176457405,
0.02965225651860237,
0.026217030361294746,
0.025261277332901955,
-0.006809374317526817,
-0.041803814470767975,
-0.015373722650110722,
-0.0019061947241425514,
-0.04106523096561432,
-0... |
SPIDER: Near-Optimal Non-Convex Optimization via Stochastic Path-Integrated Differential Estimator | https://proceedings.neurips.cc/paper_files/paper/2018/hash/1543843a4723ed2ab08e18053ae6dc5b-Abstract.html | [
"Cong Fang",
"Chris Junchi Li",
"Zhouchen Lin",
"Tong Zhang"
] | null | null | In this paper, we propose a new technique named \textit{Stochastic Path-Integrated Differential EstimatoR} (SPIDER), which can be used to track many deterministic quantities of interests with significantly reduced computational cost. Combining SPIDER with the method of normalized gradient descent, we propose SPIDER-SFO... | [] | null | 95 | 1807.01695 | title_snapshot | [
-0.02757653035223484,
-0.020286496728658676,
0.020292090252041817,
0.028782540932297707,
0.039192792028188705,
0.04726669192314148,
0.020293783396482468,
0.01849617250263691,
-0.018300535157322884,
-0.052503880113363266,
0.010756940580904484,
-0.02666868083178997,
-0.04495486244559288,
-0.... |
Generalized Zero-Shot Learning with Deep Calibration Network | https://proceedings.neurips.cc/paper_files/paper/2018/hash/1587965fb4d4b5afe8428a4a024feb0d-Abstract.html | [
"Shichen Liu",
"Mingsheng Long",
"Jianmin Wang",
"Michael I Jordan"
] | null | null | A technical challenge of deep learning is recognizing target classes without seen data. Zero-shot learning leverages semantic representations such as attributes or class prototypes to bridge source and target classes. Existing standard zero-shot learning methods may be prone to overfitting the seen data of source class... | [] | null | 96 | null | null | [
0.026438256725668907,
-0.017901789397001266,
-0.022082790732383728,
0.033706095069646835,
0.032990340143442154,
0.04864375665783882,
0.03538994491100311,
0.03271421790122986,
-0.022847706452012062,
-0.03850148618221283,
-0.0062596239149570465,
0.019864927977323532,
-0.062495648860931396,
0... |
Dual Policy Iteration | https://proceedings.neurips.cc/paper_files/paper/2018/hash/15e122e839dfdaa7ce969536f94aecf6-Abstract.html | [
"Wen Sun",
"Geoffrey J. Gordon",
"Byron Boots",
"J. Bagnell"
] | null | null | Recently, a novel class of Approximate Policy Iteration (API) algorithms have demonstrated impressive practical performance (e.g., ExIt from [1], AlphaGo-Zero from [2]). This new family of algorithms maintains, and alternately optimizes, two policies: a fast, reactive policy (e.g., a deep neural network) deployed at te... | [] | null | 97 | 1805.10755 | title_snapshot | [
-0.0627111941576004,
-0.03783059120178223,
-0.009722854942083359,
0.03298184275627136,
0.0312764011323452,
0.03631738945841789,
0.010488106869161129,
-0.0021065508481115103,
-0.021780680865049362,
-0.02627415955066681,
0.0026385781820863485,
-0.003311292501166463,
-0.06705005466938019,
-0.... |
Recurrently Controlled Recurrent Networks | https://proceedings.neurips.cc/paper_files/paper/2018/hash/16026d60ff9b54410b3435b403afd226-Abstract.html | [
"Yi Tay",
"Anh Tuan Luu",
"Siu Cheung Hui"
] | null | null | Recurrent neural networks (RNNs) such as long short-term memory and gated recurrent units are pivotal building blocks across a broad spectrum of sequence modeling problems. This paper proposes a recurrently controlled recurrent network (RCRN) for expressive and powerful sequence encoding. More concretely, the key idea ... | [] | null | 98 | 1811.09786 | title_snapshot | [
-0.01322323177009821,
-0.0098001379519701,
-0.0009212410659529269,
0.04966554418206215,
0.051062650978565216,
0.014891067519783974,
0.027676798403263092,
0.0134351197630167,
-0.024781130254268646,
0.0030254439916461706,
-0.01258886232972145,
0.00802456121891737,
-0.052651546895504,
-0.0077... |
Modelling sparsity, heterogeneity, reciprocity and community structure in temporal interaction data | https://proceedings.neurips.cc/paper_files/paper/2018/hash/160c88652d47d0be60bfbfed25111412-Abstract.html | [
"Xenia Miscouridou",
"Francois Caron",
"Yee Whye Teh"
] | null | null | We propose a novel class of network models for temporal dyadic interaction data. Our objective is to capture important features often observed in social interactions: sparsity, degree heterogeneity, community structure and reciprocity. We use mutually-exciting Hawkes processes to model the interactions between each (di... | [] | null | 99 | 1803.06070 | title_snapshot | [
0.03161575645208359,
-0.007953470572829247,
-0.016652589663863182,
0.019910553470253944,
0.03583405166864395,
0.0213617030531168,
0.0591043084859848,
0.02032095566391945,
0.004823768045753241,
-0.040502775460481644,
0.03227553516626358,
-0.0036779569927603006,
-0.07077619433403015,
0.00327... |
Decentralize and Randomize: Faster Algorithm for Wasserstein Barycenters | https://proceedings.neurips.cc/paper_files/paper/2018/hash/161882dd2d19c716819081aee2c08b98-Abstract.html | [
"Pavel Dvurechenskii",
"Darina Dvinskikh",
"Alexander Gasnikov",
"Cesar Uribe",
"Angelia Nedich"
] | null | null | We study the decentralized distributed computation of discrete approximations for the regularized Wasserstein barycenter of a finite set of continuous probability measures distributedly stored over a network. We assume there is a network of agents/machines/computers, and each agent holds a private continuous probabilit... | [] | null | 100 | 1806.03915 | title_snapshot | [
-0.017040442675352097,
-0.017327576875686646,
0.023188745602965355,
0.027620820328593254,
0.026543354615569115,
0.03680775314569473,
0.016486139968037605,
0.012833729386329651,
-0.026326969265937805,
-0.07118643820285797,
-0.006587048526853323,
-0.023768791928887367,
-0.030127502977848053,
... |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.