text stringlengths 54 548k | label stringclasses 4
values | id_ stringlengths 32 32 |
|---|---|---|
We also re-analysed available radial velocity curve. Our solution seems to be a little bit different according to the results found by {{cite:7016114b5c20db8b03532f7cb01dfb675232eee8}}. The differences should be caused due to the orbital period used in this study. Here, the adjusted period is a bit different from the o... | d | 38df20f3da5c165b8f3f53bfca650380 |
We have also checked that our calculation for topology {{formula:a03f0592-661d-464c-93e1-29b675431ddf}} agrees with the results in ref. {{cite:b29cb8c4387f23427aa30de004018298b23cd7cb}} by comparing all integrals with numerator powers of -4 (which is the highest numerator power computed in that paper). Ref. {{cite:ea5... | r | fad8e3174c5662fb165d8a8539b166fa |
Let {{formula:ce2fffcd-50f1-4ef1-8962-2ba0ea7f6bbe}} be a prime number and {{formula:62082584-0dfd-4d2f-b954-7f174a35ceb0}} be a power of {{formula:e1adf8e2-989a-483a-b935-97a886dd51ef}} . Denote by {{formula:2bec24bc-8f68-4709-8fea-90458bfebf7b}} the field of rational functions over {{formula:bedf03ed-5cf9-4892-b8b... | i | 2c153e77dc70a8a5b7438b7b159409e3 |
Our GPDs are based on the PTE effect where an electromotive force directly provides a voltage, rather than a current {{cite:60c49c49b5558bf88ed8ba0f5bbc267e21b54072}}, {{cite:b40a56f3b5baa4cf09c1e0d19fb5af3ec8105da9}}. In case of Ge, an additional TIA is needed to convert the photocurrent into a voltage for further sig... | r | 83b02d945bd0c171a9edc643c76a4ef8 |
where {{formula:df8dd0d2-8bf5-4707-930b-6788eeafab0c}} is a given function and {{formula:beeb82ed-8b83-491b-803d-03898014c8a7}} .
The purpose of this paper is to show the local existence and the uniqueness under new conditions.
The existence of solutions to the more general quasi-linear wave equations has been widely... | i | 7555a95d0e41d500701972238ddd8be5 |
A schematic of the setup is shown on fig. REF .
It enables standard single-tone microwave optomechanical measurements {{cite:fed00a2feccdbcfbcab223e4d5e38424d0b60914}}, with the additional ability to artificially tune the microwave cavity effective temperature (or population in the quantum language) thanks to a Noise G... | m | af506c99a663e51b47d0d5677362f144 |
When all the {{formula:d07004ba-c0a3-4fca-9759-623ebd0a36ba}} th order partial derivatives of {{formula:3c7394ec-29d3-41f1-afd0-bb799a4c950d}} exist,
let {{formula:7c1d04b4-9fb6-4d11-b1eb-31a52cbf3f8e}} denote the {{formula:47f8b8a1-4502-4cb6-a937-96219d9c2532}} -way tensor of {{formula:dc8627da-8eda-4a0d-9c84-dbf2d8... | r | 3eef5eca9853c8383699ce4f6de5092d |
The existing hadronic interaction models give different predictions for the shape variables. CORSIKA {{cite:8509f580a8af51283fa16b054c4e9408bd39009a}} was used to simulate proton, helium, nitrogen and iron showers with the EPOS-LHC {{cite:e0e2d16879602aca90d7455681c8b7c9532ebe7b}}, QGSJetII-04 {{cite:5d9c1e419ef103fb4a... | r | 7cc28e6c630f37a31de27998a8ffc4cd |
The proof of Lemma REF is an application of Hölder's inequality for the trace ideals (cf., e.g., {{cite:d3d18f4081d866f5d6af78a61c2d5853388f384a}}). Lemma REF then combines with Lemma REF to yield the following analogue of (REF ) for {{formula:59624dea-d44e-44a9-98da-a4b87ec11f31}} .
| r | d1e4d9df917b437b7bcf69cc9e6bddea |
For year-long QPOs, their indication for the existence of a binary SMBH system
at the center of AGN has been intriguingly discussed
(e.g., {{cite:d4ed653c40d5e87ca374ec53389b1949e320b8b0}}, {{cite:7ca0fcc45e68966ad484ad991db2b000728e9a53}}). The secondary SMBH may induce
an observable periodic signal, which reflects th... | d | b06c1a0d95a3df4a360871ccd75e5548 |
To extract the mass {{formula:b0309538-70e1-4bf2-a53d-adcf36c94a85}} ,
one can carry out the numerical analysis of sum rule (REF ),
with the aid of input parameters
{{formula:3ddad8db-879d-4465-adeb-d0a68e6a683e}} ,
{{formula:6fe8bafc-a2c8-4272-9459-c17dfaefd738}} ,
{{formula:b8b8f1c8-20b8-42b9-b02e-9c9a1b5cf9f5}} ,
{{... | d | cdbb3e3a9f3af889bb14a97790867994 |
The implementation {{formula:a627d265-afb8-4d75-823b-dafa95f6f565}} can also be improved by improving the algorithm for Riemannian-distance estimation in {{formula:236b2269-d751-4353-b56c-e08934f6e945}} . The construction of the {{formula:6e04901a-c848-4e2a-8ef9-bcdb0ded3a07}} -nearest neighbor graph {{formula:c02e099... | d | d395b90a2d47c5e91f6ac3fd76d5510e |
Seminal studies by Kroner {{cite:670e81fe6d1c937b2735464787952dd96881cb47}}, Edelen et al. {{cite:06e3ccd430ae6daa84290ec4c93e7d002f55f9c7}}, and Eringen et al. {{cite:78912a70be8d1e10092f309dbde79d26c4d4a04f}}, {{cite:1731fe6366943ee0cf0e5d299c0c2411f8cddefb}} laid the theoretical foundation of nonlocal elasticity, an... | i | dd2669f67264c0ebd2e118685e57dd06 |
A well-known neural waveform model called WaveNet-vocoder {{cite:845d16c60657a7f12afae3637191a79877d624bd}} uses a dilated convolution (CONV) network {{cite:c504c065340893ce70a0c6296985bf14383ea053}} to produce the waveform samples in an autoregressive (AR) manner, i.e., generating the current waveform sample with the ... | i | 0b69f00941d00b1b15b47d3467e0ee30 |
Furthermore, RGQD also has a strong anisotropic nonlinear response, depending on the orientation of
the EM wavefield relative to the elonged side {{cite:21a4aa32bbb3b5df2722c2ffa94a69732fc1b92c}}, {{cite:6396052cb118da86c79d31b61fa390f1772ad4d7}}. For RGQD at {{formula:82bca3e3-c6bb-4beb-834a-a1fec548db59}} and {{form... | r | bacdfeae0561d4063df43536a728724d |
In this article, we propose a two-step procedure to find the balance between pooled and stratified analysis in basket trial design. In the first step, we treat it as a clustering problem by grouping cohorts into clusters that share the similar treatment effect. In the second step, we use shrinkage estimator proposed by... | i | 69d92eed21366c6353fa39adfa70ce5b |
The observation of a significant dipole, together with the lack of significant anisotropies at small angular scales, implies that the Galactic and/or extragalactic magnetic fields have a non-negligible effect on ultrahigh-energy cosmic-ray (UHECR) trajectories. This is in fact expected in scenarios with mixed compositi... | i | 1f73ee87611b15aa6d3a672c801136f1 |
We follow the scheme depicted in Figure REF , as follows.
From the simulations we obtained RSDs as functions of time for samples
with different parameters (inclination degree and rotation speed). We
first fit these data using hyperbolic functions. Then we select four
characteristic times and use the LASSO method {{cite... | m | 3410e1e077a917ae4dbd15d2092c8532 |
In this section, we introduce our deep geometric model for motion forecasting. The model is based on Graph-WaveNet {{cite:b2ba3d0209e8125990e91fbbdfa5dd5def80c003}}, a spatio-temporal extension to the original WaveNet {{cite:46f726f98fd07f6c203153e2439118ffd4b08745}}.
{{figure:5baa7ba0-b22a-4128-ba7d-cb87141cacbe}} | m | bae80ace250d15ea8d5f0a46deb24bee |
Our template-based method leverages a deep neural architecture, which takes a single background-segmented person image as input and regresses posed and deformed surface meshes for body and clothing which match the performance in the input image (see Fig. REF ).
Before training, a 3D template of the person with separate... | m | 4a61b09ba894f468a43d4a0c679f8d91 |
Datasets: We use the full Librispeech dataset {{cite:a804e41e216e0ac59a694bcf49717a29e0a74249}} for pretraining a character based SA2SR. The dataset contains audios and transcripts for around 960 hours of audio, and we generate the proxy sentiment labels using text-to-sentiment RoBerta model trained on Twitter sentimen... | r | 37dc8ac185057bad877296a625bccdb3 |
Being social animals, human users are inevitably vulnerable to the efforts of the social bots.
Studies have shown ubiquitous social bots {{cite:7da0f27b3b121a8df49262d4f6dcbd7e87f36a56}} distort online discussions, and particularly those about politics.
During the 2010 US midterm election, primitive social bots were us... | i | c11206f252ba0e2df17178cf7216b8f0 |
Probabilistic questions.
Weighted automata and its extension as nested weighted automata, or automata
with monitor counters are all measurable functions from infinite words to
real numbers.
We consider probability distribution over infinite words, and as a finite
representation for probability spaces we consider the cl... | i | 6501d36912410cb46b81c70e5dc0b48b |
It is worth considering whether our results are affected by numerical resolution. There is a fair amount of evidence to suggest that simulations such as ours do not accurately track halo survival after they have lost a significant ({{formula:6ce24f97-c3a6-47f8-b00d-dd31349bb445}} ) amount of mass, with {{formula:c6b306... | d | cf629eb27820fe551148bdf804afd775 |
Among a wide range of algorithms, YOLO is a one-stage object detector that utilizes a real-time end-to-end approach to predict bounding boxes and class labels at once {{cite:31ce7f51995fc7ee78f4a3dfeec3c06110697259}}.
At the time of writing this paper, many different versions and variations have been introduced.
YOLOv2... | i | b7857a387d4fd7af397815d379fa596c |
We trained networks of spiking lif neurons using a supervised
learning approach which we call “SuperSpike”. This approach generalizes
the back propagation of error algorithm {{cite:7b285d21d99f43faf9fbf200163a285fbecb1f8f}}
as known from the multi-layer perceptron to deterministic spiking neurons.
Because the partial d... | m | 30b59e1a2e1c2e46fa2ab061025fff8f |
First, let us deal with the definition of the overlap (REF ) (referred to as symmetric overlap). In Fig. REF a, this local measure is shown in the case of a link connecting nodes with significantly different degrees. In such cases, for {{formula:7a03f385-12e6-465d-ba1b-f6ddcd6f8177}} , Eq. (REF ) can be simplified to {... | r | c5f4cbb719eadb43b9ea72d55106811d |
If we allow for violation of faithfulness, then {{formula:7b8e0164-d2b1-4e60-b73c-1af81cc653fc}} and {{formula:3c49f8e5-44f4-4b28-81b4-4cc814f7387b}} can still have a latent confounder. An example of this case is given in {{cite:f170a23d11a8f9ed489ad3a7f67c8bc844729e2d}}, which we discuss below.
| d | 007393b64a8fe6dfb452c6c3d3f0b1af |
We implement a fully connected neural network with the machine learning library KERAS {{cite:8115d1eb437ffaec684026047c316846ab3baf7b}} using TENSORFLOW {{cite:36a6aec40a0a5ea6d47a0cc20fb03f08240ce2db}}, {{cite:dfbb4b371c19206eb7fd18806bf56595cbd874bd}} as the computational backend. Our neural network comprises one inp... | m | edd47ab185383ece2c62addb207220dc |
Mahalanobis distance: the Mahalanobis distance (in logit / feature space) measures the distance
between a sample and class-conditional Gaussian distributions estimated from in-distribution data.
It has been used to detect OOD samples in classification task {{cite:d17a372d028329259a9f9cc749d639433f26a335}}.
In this work... | m | 004dfcd3222029322c78b4bfcab3dddb |
Loop quantum gravity {{cite:5e1a6c549fa5413478680ffd741cffb987b214e8}} suggests that, for quantum gravity in four spacetime dimensions, areas, instead of lengths, are the fundamental variables. In this approach the areas have a discrete, asymptotically equidistant, spectrum. This leads to interesting proposals for the ... | d | 45bfa33eb35b827ed448160ac5e2a0b4 |
In this paper we develop a novel efficient attention mechanism using input dependent sparsity structure in attention computation, motivated by the analysis showing the sparse nature of attention {{cite:f703d561d9854eccef2c2ee315ce9e0dcfac5614}}, {{cite:f2910bc4421edbccab47945d77e5ab0417f2e13e}}. We propose to use decis... | i | 0d1a817ce6b447bb9b8a1eaad0e6cd3d |
For approximately fifty years particle physics has dealt with a conundrum:
The electroweak hierarchy problem, the apparent unnaturaness
of low mass scalar particles, or,
why is the Brout-Englert-Higgs (BEH-boson) mass, or weak scale, small compared to e.g., the Planck scale?
This has driven much of the thematic researc... | i | e5a0e88cf7465cf25984c45cae314a05 |
High-quality whispering gallery mode (WGM) microcavities {{cite:b2db271b44738c0b9c517a093d12808eca2705f1}} have potential value in investigating fundamental physics and practical technologies such as cavity optomechanics {{cite:6a1497f0d74470735eb7eaf3967714c76b5985df}}, {{cite:40da899077a1aa490d91379cf2eb617bc96216aa}... | i | 3f06b98c0cf3a0f9fb8bfce99da55663 |
Cross-modal Fusion:
The encoded unimodal representations are fed into a cross-modal fusion network, that uses a set of attention mechanisms to capture cross-modal interactions.
The core component of this subsystem is the symmetric attention mechanism, inspired by Lu et al. {{cite:5fcf4f8166255d2406b212bf38c08e224980da1... | m | 81d22ec38b97074ac5128b58c89ded29 |
where {{formula:8a29c8f9-fb62-4dc6-9f49-11e6d335be98}} acts as a regularization strength balancing between the two objectives in Eq. (REF ) – the regularization is necessary because it is numerically difficult (or even impossible) to find a counterfactual {{formula:2899945a-8065-4b8b-a553-4bb01645b3ed}} that yields t... | m | 46160dedab7ac917dfd1e9ba24ff1a90 |
For sake of completeness, we also compared our method
with an anomaly-based detector that performs forgery localization by looking for general traces of manipulation, hence not focusing on DJPEG compression artifacts. In particular, we considered a recently proposed deep learning method, named MantraNet {{cite:4b6c8466... | m | dd9fed940377575213bf3d30796d03e7 |
Finally, we note that the discussion of holographic entanglement entropy can be framed in terms of `bit threads'
{{cite:432c7e1fda23cc3fc162d1436627312b341515c0}}. In the de Sitter setting, this leads to two distinct proposals: the monolayer {{cite:848d4930994854116d45fbdd4e344df4f8697416}}, and bilayer {{cite:61296367... | d | 9d1cbfa7f9540e662c422445829ede10 |
Energy-based Score Function To mitigate the issue of overconfident softmax probability in MSP, we propose an energy-based score function to push apart score distributions of OOD and IND samples. We first briefly review the energy theory {{cite:5e0a5ca48fcea4bd9e741f36f060ed87cb64560b}} then explain our proposed energy-... | m | d97ac741005a33eeae7aa1eadd62e447 |
In Fig. REF , we compute {{formula:ef9fc084-1ca0-4ae2-b9fb-16b9ea8c75fc}} after each iteration, where {{formula:7f57f3e6-4679-4386-a6f8-382c277da107}} is the dominant eigenvector obtained from eigendecomposition as in {{cite:49d48f6cb105f1582ac9d121da33c45a5e8fba19}}.
The MAPI iteration converges to a vector very clo... | m | 20ba9b396870e776ac87deac0a407ce8 |
For completeness, we recall the notion of relatively prime self-adjoint extensions (see {{cite:8887be605d5b4d1ec066fc4c1645994c224d67f6}}).
| r | 264449609395ea6b269f76f1b87ccf94 |
In preliminary experiments, we searched for an encoding to describe action sequences for a walking robot. The results show that using MAP-Elites to generate a diversity of sequences, then using a VAE to learn a representation leads to an encoding that can accelerate future optimizations by several orders of magnitude. ... | d | cdff6f798598d1aac8d60faf4ea5669b |
To address the choice of model dimension, we developed, building on {{cite:4428ca3d4299554fe19f6c384ddd82ba9bc8da8f}}, a practical procedure based on separate SVD of the shared covariance part and the study-specific covariance parts. The choice of the number of factors remains an important open problem. The most common... | d | 354ec424fd2e7e102320a2c39dce68e3 |
At {{formula:1b5b2798-6cd9-435c-8a71-6dc16f0dddef}} , as showed in figure REF , strips and perforated lamellae emerge as stable phases near order/disorder transition. When {{formula:8964b18f-beaf-4455-8bdd-fda120015b21}} , NCSLM is observed. NCSLM is an ABCBA lamellar structure, though the width of the two B layers is ... | d | bfd1593e7881d576aacbbff1a582fbc3 |
Predicted segmentation masks were post-processed by applying 3D morphological closing and extraction of the three largest connected components. To demonstrate the utility of our loss penalty term, we compare it to other successful methods in Figure and Figure , using error maps and the following metrics: global Dice s... | r | 21b3b7d1ea85332aad45b59964eeabc1 |
Human feedback for judging simplification quality is more consistent for sentences, compared to longer samples, such as entire documents.
Metrics such as BLEU {{cite:4fa073d52819bf852a72c216517cc902ffab4d21}} or SARI {{cite:56814097c4ba36b8de107e82f74209f8ba39a10a}} rely on (aligned) reference texts for automated eva... | i | d0f33ee3606734813ecda52b96cc5de7 |
In this paper, we explore the mechanism to enhance the anomalous current caused by a background magnetic field in the spacetime with a boundary. Usually, the anomalous current is suppressed by the mass and the distance to the boundary, which are the main experimental obstructions. Remarkably, we find that the high temp... | d | 9a25549d8dabab265f8b307948c63367 |
We first describe the scan method in section . The results for the global fits
are presented in section . We begin with a comparison with CKMfitter and UT{{formula:1cab52d7-dfac-40f2-9764-e77146602604}} ,
employing the scan method to extract CKM parameters using inputs standardized for the book Physics of the {{formula... | i | bb61020a1a93bf4bf0890d00966a08c0 |
In {{cite:5202d57aeff390599b8980b5f8cde871cf5f5a5c}}, it was shown that a stacked self-attention mechanism could achieve state-of-the-art results in natural language processing tasks {{cite:323f6f57a6db82b8d33ab3a5d364c153c3119757}}, {{cite:bd66ec3be43cd6ded8daea9d8c841489ff054f2e}}, {{cite:d3568b93e8087142ae32662f687a... | i | 998eab98eabfbac9ccb67c55586b1930 |
The Alfvén waves lead to non-linear interactions which are crucial for the dynamical evolution of a Kolmogorov-like MHD spectrum {{cite:cd4efa0a858fb8eaedfc8372fd42800fb89e5002}}. We have also performed a power spectral density (PSD) analysis for all IMF vector components. It depicts Kolmogorov-like turbulent nature (... | d | 9e2c92853914c22426b1058e77b8b1fa |
Our method uses weight sharing, that is, all fake-quantized weights and activations are obtained from a single float tensor, and are generated just once per-layer on the fly, i.e., we have {{formula:ea338162-2e2a-4b86-ab38-409b42c24317}} and {{formula:2aac75d0-5667-47cd-bbff-1b243389cba5}} temporary copies of {{formu... | m | 7c350330863ef93f265481bdceb9e1e0 |
Distinguishing between the near-field and far-field is becoming increasingly relevant as modern wireless devices begin to operate in both propagation regions.
The most common near-field distance, known as the Fraunhofer distance, is
proportional to the square of the aperture and inversely proportional to the wavelength... | i | bd27ee43daa8b0e3b79fe3e9ca26efa4 |
To summarize: given this state of the art from prior work, it is clear that alignment-based approaches can outperform mean-based algorithms at low deletion rates, but it is not clear whether, or how far, alignment-based approaches can be extended beyond the {{cite:c99680edd22ce74c2621e38c30a271937d6bf430}} results. Fur... | i | 240a9adba592013a4606b1edefb17832 |
We present two models for the multiwavelength emission of Tycho, which both adequately fit the SED.
Model I allows for the weakest magnetic field in the immediate downstream region of the shock, {{formula:0774de40-4fce-448f-9c83-839cc26075e3}} , that is compatible with the entire {{formula:d238b5cf-fbb9-4612-ad46-9119a... | r | 1a717d94ecf034be795601f213538ac0 |
In our experiments, we use a different iteration formula based on {{formula:d05b07ff-231d-40cc-96e6-cb1f40652e1a}} -normalized gradients {{cite:f0d7640bf6d7c92dc0d99beeb9cf794f840e743d}},
{{formula:2fbb985a-cd40-40eb-a282-1d1310a6f33f}}
| m | 4b2d93756dbab6998a2af6499aead55e |
In this section, we present simulation results to verify the performance of our proposed distributed EM algorithm. In our simulations, we use the pairwise gossip scheme {{cite:16895ab7da7ce07579eb3ec6c538207df0b06e86}} in the Gossip Step of the distributed EM algorithm. Specifically, at each iteration, two neighboring ... | r | 07f2ffd4da2ee30172b5e64783ce88c0 |
Thus, these findings provide the experimental realisation of fractional optical media. The results offer applications to engineering optical pulses and their use in data-transmission schemes. Further, implementing the FSE in the temporal domain makes it natural to include nonlinear materials, which is a promising direc... | d | 2789ff373139ea0d9df0d1b38b09be8a |
[Ungapped hardness of Set-Cover from {{formula:bd500b83-e293-474c-9e3c-3ed525d8192f}} {{cite:048c2b2461faa935e94517e19adc6e987df696c4}}]
Assuming {{formula:8bef45c8-9699-41bc-91b8-6882d975df7d}} , for all constants {{formula:699a9d3b-a60f-4882-8f8d-2039dbff15d8}} and for all {{formula:1a12cf2a-5ac4-4200-841b-b0a3f6a... | r | b914fe98f2b75f4ca542fde90153ef04 |
It is believed that the ringdown waveform is dominated by the QNMs
of the compact object remnant. Thus, the detection of overtones from the ringdown signal allows for precision measurements of the characteristic parameters of compact objects like the mass, charge and angular momentum. Various studies suggest that the r... | i | 73890ed30b2b1449922db56e6135e174 |
Recently, a new graph-based classifier proposed by Papa et al. {{cite:f7469139d78a0d6094a6d92a5f4eec4d20f94980}}, known as Optimum-Path Forest (OPF), attempts to fulfill the literature with a parameterless classifier, which is effective during the learning step and efficiently when performing new predictions. Several w... | i | 630b5e1dce03230f8327b66d464f5292 |
To deal with the absence of keypoints (joints) due to occlusion, truncation and low-resolution in real-world pose estimation (as shown in Fig. REF ), we construct a multi-expert pose estimation system to predict improved 2D poses. It is fine-tuned with the pseudo ground-truth 2D pose generated by a novel Selective Spat... | i | 4e3ae1ec3786aecbc783c84d5b1b3e50 |
Our formulation allows borrowing of information across {{formula:abcf781d-199a-48d8-9af6-45c2a5d308d2}} and {{formula:7aa72c0e-8240-409d-ad7b-e28cb3619af0}} only when the corresponding actions are same, through an indicator function. A traditional nonparametric estimator will borrow information even if the actions ar... | m | 13b14c67884cffb596ecacd1c4b17ca6 |
Similarly, there are several variants of SG-MCMC methods that have not been benchmarked against DGPs. Havasi et al. {{cite:600e882b2b5fbcc5b6aba90b5faeb3486254d609}} only considered the original SGMCMC {{cite:5a61d336c47e9c672f04c6d2f3be9f001c3b5a45}} approach. However, numerous variants of the method have been introdu... | d | 08d372f52d87919053909c7ec6365e30 |
Crucially, in the context of random networks of large width, LN's operation at each layer {{formula:3038c464-7fd7-4da6-9a11-23371b24c5cf}} does not compensate this mean shift. This stems from the fact that LN's mean and variance statistics can be approximated by zero and a constant value independent of {{formula:78b18... | d | 0477f47677c38fb1d9f4cab8dcbbff56 |
Fig. REF and REF show that our GVS fills pathological regions with diverse healthy-looking tissues, which implies the reduction of the mode collapse. The research {{cite:68013a347e11d56b90753a9f545530636fcb52ec}}, {{cite:e1490900d555759465f8a811b49d9e9f12b491ee}} demonstrated that the higher dimension will exacerbate... | d | f9f00f2824ee5c2924a98d85a01e2ac2 |
For the MIMIC dataset, we can see the models' performances in terms of both AUROC and AUPRC in Table REF .
In addition to the models we previously described, we also compare against the interpretable medical baselines of SOFA and SAPS II which are simple logic-based scoring methods {{cite:af0e4387d6e416682f5ef8f0464aac... | r | 557955de096f66aa8624d5b209817877 |
The above assumption measures the population diversity (heterogeneity) between the gradients. In heterogeneous settings, this bound indicates the similarity between different distributions. {{cite:25ef518355e5f9a9c538d43132acdd5742a95743}} show connections between heterogeneity and the Wasserstein distance between th... | r | 3bd2037a236b4e15991f11cce51d0dc6 |
The small effect of the training dataset sex composition observed in our study is in line with previous research showing sex-related differences in MRI recordings to be limited and gradual {{cite:23b16ddf268619a2f2890ed6e29435524a03eec7}}, as opposed to the apparent sex differences in chest x-ray recordings.
Neverthele... | d | 624b26d455e90c736263dd29ba0e0640 |
The combination of all three techniques is applied in our experiments with 27 languages involved with the Transformer based model {{cite:7602ebada3f061d0ef437fcddf71c5b980c1259c}}, {{cite:d8e70d1f61c6a0f11ce06c87371a055f1ab416aa}}. In the process of learning these languages in different iterations, we were able to mini... | i | 89960a4175488e1854125589b35a1a88 |
Despite the pixel-level bit implementation is obvious by looking at the bitrate distribution result, the reconstructed image is less obvious. It might have something to do with the autoregressive context of {{cite:23680a8499171bc652fdfc6b1a049637e52a6a39}}. The autoregressive context of {{cite:23680a8499171bc652fdfc6b1... | d | 05fc58289de18ceee9582803711bd21b |
Despite the wide applications of DCNNs in modern learning tasks, theoretical understanding of their generalization ability has been challenging.
Promising progress has been made recently.
An estimate of approximation error of DCNNs is given in {{cite:65753742c6c07634d67f4b8c0ca2dc375ebac7d2}}.
It shows that the approxi... | d | c6af83c47460085e1c9d8438519c81ed |
To prove Theorem REF , we note that any finite set of time-frequency operators parametrized by a collection of finite points on an integer lattice in {{formula:4e8749fa-81ad-481b-9d23-5dfbdc08ca3a}} , forms a collection of pairwise commuting operators. As such, the Spectral Theorem {{cite:438f78d1493a649180f42b38be8f28... | d | 9d8c582f4b0c3d34e7db175d8348636e |
In this section, we experimentally prove that ADC is more suitable to human pose estimation than deformable convolution (DC) {{cite:b196533919fdfcd73e764151974dd580debd9ca6}} and scale-adaptive convolution (SAC) {{cite:0d193dd138230a5df8cb17a139b50d19c5886865}}. Comparative experiments are performed on SimpleBaseline... | m | fdea12781231481e86cb0d6e17facd9f |
Open-domain dialogue response generation (RG) models aim to provide human-like natural language responses given dialogue histories {{cite:9d7f1dc5a258d40975b5003db0999b3ef1d4fddc}}.
To improve generated response quality, many studies have been conducted to develop knowledge-grounded RG {{cite:2edfaacebc964a47f6f68c69d0... | i | 90cf3c641708c65391f8170824fb468e |
Decision trees {{cite:b375f0a2c4ae02e94ef66140ba9e5e281ec36ae5}} (DTs) are generally considered to be an intrinsically interpretable model family {{cite:69d3d2f73b58c3cfe762113fbf5a429da0a83df0}}: sufficiently small trees can be contemplated by a person at once (simulatability), have subparts that can be intuitively ex... | i | 20902d2f127ef211e589a059270a922a |
Scalar fields are ubiquitous in modern particle physics and gravity models, including inflation theories and dark energy scenarios {{cite:dd8199f453f3929991fcde774105c2c8b422a998}}, {{cite:4474c32bb16e5d3a79987d92643845d4428a5618}}, {{cite:ca8ab4cc94deff8a70ce3b8502c5b0510e269344}}, {{cite:6fc67eebb73c50963db33d8f8f519... | i | bac9c249ab544e50d156b166ca6e6e0f |
Driven by the rapid development of computation hardwares, the scale of deep neural networks (DNNs) is unstoppably increasing, where the amount of parameters has significantly exceeded the sample size by even thousands of times {{cite:c702464e9ce852b60e996fccad411e64f0da2652}}, {{cite:c4b4bea017f551331796d7a324825352262... | i | decb0b88d22240b8d626c4d302930f1c |
See {{cite:7323949ea7f62ddba20b6cc93c82658c5111a52a}} for more details.
| r | cd26dbf9fe36f99d134662ccdc87ecf4 |
We will also use the following lemma from {{cite:b974da5c99445c46b25c260c9063d32162edb6c6}}.
| r | 1743304c74db1f2f5bfd9175399e0d98 |
To establish that {{formula:a8ab5127-81ed-4765-b50e-c5596dde896d}} is GC, we will follow the methodologies developed in Section {{formula:fec0ca55-2162-4fb1-a751-8265f011f0ae}} of van der Vaart and Wellner {{formula:12a79375-9843-4181-b56a-dd5e992c3441}} {{cite:34b5fccb4fea991606dd7f65223f2eea3c37dd0c}} and Kosorok ... | r | f3953a261f29d7821f95c6b17d9f3719 |
Denote {{formula:3daa116b-6ef7-42e1-9f34-13f9bf692f33}} as the set including all the accumulation points of sequence {{formula:7a7c193a-7b14-4938-a1c5-1afa01921009}} . It follows from {{cite:aeb647202fb2c38b60c90a12d07dbd516cf5967d}} that
{{formula:a4db30bc-b03c-44dc-af65-83796ea9ec42}}
| m | 237d5143042e493294c18e1094821bba |
The characterization of {{formula:9c22164b-718f-45b4-9505-85f8b98660f6}} is distinct from the characterization using instrumental variables-type moment conditions in {{cite:09713f46bfdc8da67226658474d840992accd7ad}} and implicit in {{cite:6b4fe93f38301856fdc124d648b62a5b0f79b799}}. As we discuss below, when our assump... | d | ff0f76f59a2048990358a54316368c98 |
Many calibration approaches disregard the role of refinement, leading to severe loss in the utility of DNNs thus trained. We successfully presented the case of declining refinement for a wide variety of approaches tested on many different datasets. The derived relationship in equation REF showed how improving refineme... | d | b5b5160d7eab06f7b9087453d503e4e0 |
This phenomenon of regularization by noise is now well-studied in several situations. A lot of work has been devoted to the additive case {{formula:0b0748ba-b171-498c-bfd3-98a0cbb9ef82}} and for several kind of processes. One can think of the seminal work of {{cite:240930552a5f6a1fe7cacf1b23c82748bcd2aa9a}}, {{cite:27... | i | 0d1a83978a2c58d9762c5dee31e61d29 |
Quantitative comparison results. We compare our method with three existing semi-supervised learning (SSL) methods: {{formula:d30b3762-e640-497e-bfd3-52cdd704b3eb}} -model {{cite:8871c8e6b7e9ef90f3f44f62bc1ecba67bc45603}}, temporal ensembling {{cite:8871c8e6b7e9ef90f3f44f62bc1ecba67bc45603}}, and mean teacher {{cite:50f... | r | 9103a8f70f5376ae3be0a6891d011a86 |
Table REF demonstrates the results of wrong operator localization and repair. We report two accuracy metrics: localization accuracy and joint accuracy of localization and repair. We can find that compared with graph-structured baselines, Tree-Transformer achieves a significantly better performance, especially on the j... | r | f9e6d085d609b9ea66dcf69935920584 |
Furthermore , we can also exclude models of distant dusty absorbers that diminish the
UV by extinction and simultaneously diminish the X-rays by absorption.
However, dust-free absorption, arising in the accretion-disk region for instance
in the form of a clumpy wind, is an efficient mechanism to cause X-ray spectral an... | d | e12322d0befc86910b6da4d3bebd27d6 |
Remark. Note that, in the formulation of constructing rank dataset, the selected perceptual metric is not constrained to a single one but can be any combination of multiple metrics. For example, perceptual index (PI) {{cite:677636954e7a563dbc3d27f15b702505432d22d3}} is a combination of NIQE {{cite:b8042971e79e6b4d59d48... | r | d728d1de4214d8750f3556f488458a68 |
The basic strategy of the proof of Theorem REF is to apply the fundamental ideas in {{cite:f6a358f84a46316b5c1929d38118f466ccbd2691}} to the case of group extensions, by using localization algebras, twisted Roe algebras, and a geometric version of
infinite dimensional Bott periodicity. We use the coarse embedding of t... | i | 0b82df143991c73520f59ee406585240 |
Therefore, one should evaluate the kernel centered in each training sample. If one relates the proposed LDU model to kernel ridge regression, the main distinction is that we do not evaluate it across the entire training set, but across a smaller set composed of the prototypes. Hence
{{formula:b1e202b1-1c59-4eaa-a2db-d5... | m | 954f8623028d7877fb2bc07fbb74c4a3 |
To calculate the {{formula:44db7ee7-dc7b-44d5-8f9f-54e5997cc67e}} for stratified models such as this, one must calculate the so called next generation matrix {{cite:f462293921d7a0ea46c111f2170ec3e66ae8c906}} which has elements given by
{{formula:e7b46ac6-3ec9-4993-96d4-72d9fa1d8778}}
| r | 478c52742796a7c1f1b5a75496abddbf |
Neural machine translation {{cite:dbe3f57cf4660a72cbdc12e363c75f25f263d12e}}, {{cite:979e28ad02944eb2c0b36a5b0d625c425efefe19}} has established itself as the new state of the art at recent shared translation tasks {{cite:4ce954a9ad7f78d6ecfe0ea6bd66d721bd646d02}}, {{cite:eb258ec21899e7e09659970c9705a6cd8d81eda9}}.
In o... | i | 4ee877cc0d458c3f7fccf73e15da98fa |
We have evaluated on both commonly used distance metrics: Euclidean ({{formula:5b1728c1-b141-46cc-b829-b40961da47aa}} ) and {{formula:ed359f10-349c-442b-ad9e-c0efe3eb13d1}} {{cite:bdef15fa31e448a75c3f9ef46169a1b9c6584aa2}}. Given the subtle differences or the continuous nature of disease progression {{formula:087478f2... | m | 9fb743a21dbd2fcee063c6fb1e3cadac |
Finally, Figure REF shows test accuracy achieved by our method as compared to the PGD approach (for {{formula:494e6322-a456-46ce-9ff9-219fe5a7b49e}} robustness) and adversarial smoothing (for {{formula:26028de9-0c19-4519-92e4-63faf8b56d99}} robustness) at various intervals (wall clock time) during the training proce... | r | 97c9d1247ac7bd4ca6307fc0f0e77aa9 |
Let {{formula:68004931-dd77-41ed-ae9c-d1da6592ec65}} with {{formula:f5a2200d-46d9-42c8-bce1-574ea90a8cf7}} . To prove that all points in {{formula:34d74ee6-515f-4c78-b7dd-43697f25dcf7}} have multiplicity one, consider the Jacobian matrix
{{formula:241929d0-0d1f-4e78-85bb-de83305d3ca2}}
The point {{formula:e10341a1-c... | r | 2b8ba001304512770f495722add5e15b |
In the experiments, an episode terminates when the agents receives a non-zero reward or the maximum number of steps allowed in one episode is exceeded. This maximum number of steps per episode is 1000, whereas the maximum number of total training steps is 500,000.
Furthermore, we set the Jensen-Shannon divergence thres... | r | 6a70163e5372f49f9b09fbd4c6994b9f |
More specifically, we refer to knowledge distillation {{cite:e3c2a7c37eb007febecd90aa5d198ab27cb0eb9b}}, i.e. the process of transferring knowledge from a large neural network to a smaller one that may be appropriate for deployment on a device with limited computational resources. Another area of interest includes the ... | i | 0aaf0908d2ede587ee3b7ede5b2d0407 |
Since our method is in principle applicable to any gated RNNs, we further validate the fast gate on another popular gated RNN, GRU {{cite:8170bfa1909b6c72ac1cd90f8da5411c22d1ab95}}.
Additionally, we evaluate the fast gate combined with the gate initialization technique, called uniform initialization {{cite:6abe7f531000... | m | 680df8450dfb2cd1887d15b8985dcfb3 |
Single crystals were grown by chemical vapor deposition {{cite:f9a2ca3a8fde356b3af8e252a3d6f0a40ac5a85e}}, {{cite:dacda3a67cda024ab010e2a3a11d6922e401f512}}. Their orientation was checked at room temperature using a Laue diffractometer. Oriented plates were prepared using an electrical-spark cutter. 15-{{formula:59f394... | m | 11f2e0af324d4ec2f655f9636972039b |
We show more results on a variety of model architectures.
In particular, we choose two CNN-based models RepVGG {{cite:bc5c0d5bb7996d0d7e16abc57cfa9e903032cf26}} and ResNet-50d {{cite:605a423ce945d7ee92058d0e85bba3e612a2db7f}} and two transformer-based
models Swin Transformer {{cite:19667ecba2d7384dc87a3689f4195be341db7... | r | 033032b0ec1569d30eb746513e6444b3 |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.