text stringlengths 54 548k | label stringclasses 4
values | id_ stringlengths 32 32 |
|---|---|---|
Retrospective simulation {{cite:dc7ce3ec80fc5a2f19f14284b7f8664f770bcbab}}
is an attempt to take advantage of the redundancy inherent in modern simulation algorithms
(particularly MCMC, rejection sampling) by subverting the traditional order of algorithm steps.
It is connected to demarginalisation and pseudo-marginal {... | m | a0301468d28ebb672d39254bdfce79a9 |
Despite the loss of asymptotic flatness for a magnetized spacetime containing a black hole, this solution has been considered to have some astrophysical uses especially in modeling the spacetime near rotating supermassive black hole surrounded by hot moving plasma {{cite:7047d371c50cc1dc347ee88d3ee04f19b8563df8}}. Inde... | i | eae8b866f4f519c1377d8872a0ede18e |
In the SME EFT formalism {{cite:0cb2b40ccca94df10434fc60552e8cf3fcd05ee1}}, {{cite:c8df0547d54a6689a46dc915a8994433454fd273}}, {{cite:ae9f4f1891eee3d1f3814317dce6909887d62f86}}, if the apparent cutoff above {{formula:88dc9454-8df1-492d-8347-9b1e22ae2a4b}} PeV in the neutrino spectrum shown in Figure REF is caused by ... | r | bd148abc8d67b6bb58a92d006c0e8208 |
On the other hand, for a few known classes of games, {{formula:1712e650-7de7-4d94-a06d-239527c87280}} is computable, and sometimes computing {{formula:69494ed8-5796-485f-96e6-231d5a775457}} is even easier than computing {{formula:e3670a6d-372c-409f-a1d6-72fd57462c15}} . Cleve et al. in {{cite:d63ecefc2e452f494cd49c33... | i | d7fb702d3e62db0d0f91421eddf8c16b |
Along with the proof, we have extended standard initialization schemes to non-zero biases and formally shown that our proposal defines well trainable neural networks, while they support the existence of strong lottery tickets.
These initialization schemes include the 'looks-linear' approach {{cite:33998fd05626cb23267eb... | d | 3fe3eda0acea5d696c5d6b1110123d49 |
The above formulation has been shown to be effective in traditional offline hashing methods {{cite:acb4a83705605ca81a4ccce4a42b66789425ed75}}, {{cite:337d8f213576d79addb1e3a11f528c81ad079469}}, {{cite:e01d0bee83539aaf16ae9e1115e81d4cb51ec035}}, {{cite:c7639ff15a45a451c8ab0637744cb573611a7ee6}}.
However, directly applyi... | m | 1f7dad5c527c0688926f025be2f75e2f |
The memristive reservoir of the network type is an attractive option for hardware implementation, because the number of possible network structures, capable of producing different dynamic responses to input signals, can be drastically increased by scaling up the network size. This high scalability is the major advantag... | d | b700f7cbea901f87300561a7c1b286d2 |
Hyper-parameter optimization (HO) is the process of finding a best set of hyper-parameter values that cannot be learned using the training data alone {{cite:62cd3d349117805c40614360e33824b64466088a}}. An HO problem can be formulated as a bilevel optimization problem; the outer objective {{formula:e4cd39a7-baae-48b1-901... | r | cf68395f2cc4706fbcd62635a79e9f5d |
In this paper, we separately described the travel mode, destination, and route choice behaviors of travelers using the free utility model. The free utility model has two basic assumptions: the traveler pursues maximum utility and needs to pay information-processing cost to master more knowledge about the utility of alt... | d | d80038e5f032eb75b0bf3e5049a064b5 |
While in this specific implementation we used random sampling and nonlinear least square optimization, our framework is flexible enough to support alternate sampling and optimization strategies with the underlying factor graph acting as the core data structure tying them together. As such it should be possible to use g... | d | fe3a16123564190824f87660533d7466 |
As a remark, in this paper we write down fields in MUR of the Lorentz group. This approach turns out to be more natural, compared to the approach that employs Lorentz indices, towards the goal of classifying all possible vertices {{cite:1b2bad1260a3402cd3f57909ccef65a7f9f18254}}, {{cite:c3977b024b273365d98883fc429e44a0... | d | 1af1fbeb46c427214c55555e4483a2cc |
[leftmargin=*]
GRU4Rec {{cite:eb9221914d45d2476af355d4022861082a0c1bba}} is a GRU-based session-based recommendation model which also utilizes a session-parallel mini-batch training process and adopts ranking-based loss functions to model user sequences.
NARM {{cite:d5c0eab63de9958404e16ff30d6582a9e799af70}} is an RNN... | m | 1bea8f47e628777c4da7132da247d31a |
Intuitively, the idea of using the concept-feature dictionary to the ViT training could be similar to the memory bank mechanism in MoCo {{cite:3d2e6a8659d7a0fd33733567efa18e20ee3f4709}}, where the features are stored in a queue for replaying later. However, the difference is also clear: we have multiple queues that are... | d | de6ae8ed4fdd436f41cda024c5133c80 |
To illustrate the benefits of leveraging multi-step information, we compare {{formula:38bb7941-8b12-4584-b431-6b663ae18c85}} to the one-step estimator {{formula:14809fc5-e0a5-4999-8eb0-9447edad2b06}} and also with entropy production rate
estimates from the thermodynamic uncertainty relation (TUR) {{cite:2630c1ed678de... | r | 2b930a698856f29e085f02c7bd54c61f |
In order to observe phase separation between the two components it is necessary to include fluid-fluid interactions {{cite:3e0f6760ab1c5e048cd78a0d16388cde2db504d8}}. We employ the model proposed by Shan and Chen (SC) {{cite:63ef8ceb060c03e00800cf3e9c13263858b36cb3}}, {{cite:d0f942f9cbe3ff8645bd72babe34498872e05001}}, ... | m | 23865722445de0864091a5fefdb9b956 |
Table REF summarizes quantitative results on PASCAL-5{{formula:dfc61c12-c1ad-47da-a515-fb8adc9b32d5}} {{cite:cda1c6342ace693c6f0fd5c31d649f901d26ccac}}. The tests were conducted on two backbone networks, ResNet50 and ResNet101 {{cite:c9f019a819f14b07227539af0c3d5d1ade57b7df}}.
The proposed method outperforms the othe... | r | 3c2764191dff0710e6443d50d7052f93 |
A critical task to evaluate the quality of embeddings is link prediction {{cite:3ea4dd4d0188672b9657ff62a2c1f258c4eb9b98}}, {{cite:545d220656ef0ed8d7d50a8591d558b5d2877d5d}}, {{cite:4b6cacb55b7209d9324fc588ee9c615f6b6ce1e0}}, {{cite:41b7fbbd39741d122b6f17c782180e2eb3677388}}, which tests the capability of embeddings to... | i | c00046c67c78fce43f2df44d8037aef4 |
We expect this nontrivial coupling of soft and hard sectors to
generalize
to nonabelian gauge theories and to general relativity, and to the
quantum
as well as classical regimesAs discussed in Sec. REF , the statement
that the quantum theory does not admit
a decomposition into two decoupled sectors is not in conflict w... | d | 6f5f005208e9993c2f9e2b61eb6cecf7 |
The main result of the present work is a refinement of Theorem REF into a LDP both for a Nash equilibrium and the state processes at equilibrium.
It is well-known by the celebrated Varadhan-Bryc equivalence, given in {{cite:0e6e8120221290889d2a101fe36ca0581611f578}}, {{cite:d6eac622b64fbb997b5ce883f6845603d9f046d5}} (... | r | 8f2e071545a8e2d49260e323099e817a |
To train the RL agent, we employ a deep Q-learning approach with a replay buffer {{cite:c74ecbf02def13a5c41a620250463fa5e4f34019}}, {{cite:8f482aaa2be03be3838116bf7f05723bf2477d7c}}, {{cite:85f383ce3fdda825dcb9789d39c0d5bf6d118b39}}. The deep Q-network is fully connected and contains 4 hidden layers, each with 64 nodes... | m | 925ab556eced4bdcc62a95de1a820ca1 |
Thus, regardless of the sign of {{formula:9f3e73b0-e18a-4de6-9a68-9a70ad192c8a}} , the band gap opens at the {{formula:6e69f243-a097-4e3a-9050-61d4b21ab044}} point and this gap becomes
topological. This is related with the fact that the irreducible representations (irreps) at the {{formula:0b130a4a-3092-44e5-a898-2e57... | d | 46b1d1bb57b9422a45ad3356e237f5da |
In Table REF , we show the reproduced results of the original
QDE method following the exact experimental setting as implemented in
{{cite:7028ee0e7c61fb02c7118b5586aac7697c87f3d0}} pearce:2018. These results are included to
offer a comparison between the 1-layer and 2-layer models (Table
REF and Table REF , respectiv... | r | a424ef8fed0f7fe79abf4e6576796e8b |
Efficiency. From Table REF , we observe a significant improvement in inference efficiency between our TCSA-SNN and vanilla models. On Gesture, three-layer TCSA-SNN (following {{cite:1a61a6712cc7deaefd6465a2b8837490c9c525f3}}) and five-layer TCSA-SNN (following {{cite:2e532b9ac6e987aebe9a626e635c3f7c63b760d1}}) increase... | r | 451aa233d8ac38a658986eaddbd12521 |
Information novelty might help explain why politically-opposite feeds increase engagement. The fact that respondents indicate politically-opposite feeds as different from what they are used to seeing in their own feeds implies these opposing feeds contained novel content. Prior studies have demonstrated increased user ... | d | 8ff8815e99344bba6710d4e9a37c4cc5 |
Meaningful Perturbation (MP) {{cite:8c89231cf4a24cd2e480a1448ace646300d56b01}} finds the smallest real-valued, Gaussian-blurred mask such that when applied to the input image it minimizes the target confidence score.
MP is sensitive to changes in hyperparameter values {{cite:eb208579d820ae88903ca5e3d5e8eb5c87985c14}}.
... | m | 6fc05635db582296f757a4d3ec3f3012 |
The {{formula:7220bf3d-0b7a-4d76-aa2f-24590dd694e4}} is used as an input to a highway network in order to reduce the interferences from the rest sources. The encoder, the decoder, and the highway network are optimized using {{formula:d86e1890-a685-4f5d-b8a0-b455a8616357}} between the true magnitude spectrogram of the... | m | a567c6e9946dbc8050775e775114682f |
Diffusion is a common phenomenon in nature and associated with a variety of processes, resulting in inaccuracy and misleading of the underlying physical phenomena {{cite:00a32aafbc2552522c63aa09806c86db829a2b68}}, {{cite:8daeb7f7c895dee57011a1322560d16727bbd7dd}}. In this work, the term “diffusion" will be referred to ... | i | ec958584fe59a68cbebe33a04e7e394d |
where {{formula:f41d43bd-bedc-4347-8bd5-dc983b6831cd}} for Max-cut and {{formula:d59ac25f-31e8-4ec6-921e-edf93ebc94ad}} is randomly chosen to be {{formula:80adabc5-463c-4377-81e7-a34c4ff96c95}} for S-K model. For graph coloring problem, we first model it with quadratic unconstrained binary optimization (QUBO) form a... | r | 1d3f034f3506412592a0a724ec092150 |
Embeddings from all layers.
An alternative approach, inspired by {{cite:2be745d0a6918b0952b9ae893d81a06c56ee02c4}}, is to use pooled embeddings from each layer (not only the last) of the Transformer. We explored two pooling strategies: averaging over the timescale (common for speech Transformers) and taking only the fi... | m | 10aff5b49bd54643d200772d8a7d0b76 |
Complementary information to the above analysis comes from the
knowledge of the electromagnetic form factors, assuming that one
of the massive constituents of the bound state is charged. Their
asymptotic behavior qualitatively probes the compositeness of the
states: a rapid decrease would be the signature of a many-bod... | i | 4c69d46b2d91b8990a7e3cc6d64aba95 |
The network of gravitational wave (GW) detectors composed by the two LIGO {{cite:e6a267030d1f8e607254ecd5ea21fa704f7a0108}} and Virgo {{cite:4715525a70922b88fd334ec1936bf5054235f7ba}} observatories
has already completed three successful observation runs
{{cite:5beda308cbd3407ea67aa34a11c4e390936bcb2b}}, with the detect... | i | 195dc297fd59d2d4efe4cc8cfe6290f1 |
Throughout this analysis, we have focused purely on linear methods.
In recent years, nonlinear methods for dimensionality reduction, such as autoencoders and diffusion maps, have gained popularity {{cite:ee4e697cb444791abe553f034699fbc4029da962}}, {{cite:132446e66b9ee02f916af965d18e81b3a149bf58}}, {{cite:a0314dc52c187a... | d | e32947ef344ae03dafbb015c0a756172 |
On the positive side, we show that, when {{formula:bdef11ef-ce7d-4163-ba0d-65c60dec858d}} , the optimization function is submodular and thus can be greedily approximated within a factor {{formula:a57848c3-03a1-4e78-bb65-2d8d0029cb6f}} .
Preliminaries
Structured populations.
Following the literature standard, we represe... | i | 2fac653b720e38f94be501860611a156 |
After the trial runs, the results show that the overall mAP for all of the models and settings is between 68% and 81%. Table REF shows the complete results of the experimentation process. The overall best classification was the ResNet 50 architecture for all of the settings of the superpixel masks. The three variation... | r | 3f911bce4f9aff73e734cb2fc5772d75 |
The architecture of our model that maps the fingerprint to its location consists of a CNN inspired by AlexNet {{cite:4ed3e16c2c93e15274c30234411aa72a6671fde1}} and ConFi {{cite:a9511dff4b03362b9ea8a2cc7aeccb2a9e641dc8}} (see Fig. REF and Tab. REF ). The fingerprint – a CSI-image – is fed to the model which performs a ... | m | 5c60130fefc45daca25a317dd503c0f8 |
In the context of gradient flows for diffusive systems, recent progress has been made in developing energy-stable methods through a variable transformation representing the energy, generally a nonlinear function of the state, as a quadratic form—such methods are referred to as invariant energy quadratisation (IEQ) appr... | i | 8978131cf0a939eac5f0f76ff9dbce4f |
Similarly, we train and evaluate a CNN based network TIRG {{cite:71f82e9b8979d9b959e495c7f3f0c3aa38de0873}} on GCD. It consists of resnet and LSTM backbone for feature extraction and {{formula:c5eb2827-5329-4fde-a411-1f31156dabce}} layer for feature aggregation. Herer {{formula:618e1aeb-f107-4765-b5a1-e75ab530fa01}} ... | r | b76b8cd30717f2adc41ac12dfdd524de |
In this section, we present two technical lemmas which capture the
essence of the proof technique in {{cite:6c56b3cbfcc525a05930529d41f49ec667b78f21}}.
These lemmas are used in the proofs of Theorems REF
and REF .
Though the present author discovered these two lemmas independently,
it appears that Lemma REF is to be ... | m | f7e4bcb09c06d96a0c4eac90ce291b71 |
While DL models are able to make predictions with high accuracy, their deterministic nature prevents them from providing uncertainty estimates for their predictions. Wildfire occurrence, however, presents a high degree of stochasticity, therefore uncertainty assessment is critical in decision making. Hence, Bayesian ML... | d | 7b54f75307e9868edf0cb03435d55c79 |
Next, Fig. REF compares the rate performance of the proposed scheme to other counterparts under imperfect CSI, where we have {{formula:2b9c1634-dcae-4ba7-94ed-10ca7f05e7cb}} . In our simulation, we consider two benchmark schemes: 1) the AO method characterizing the near-optimal performance {{cite:3ab4c0b7d071f18c76dde... | r | fe9d901d959734dd93233bb5d477ff56 |
Full Fact's approach centred around using sentence embeddings as a feature engineering step, followed by a simple classifier such as logistic regression, which is what we used. They used Facebook's sentence embeddings, InferSent {{cite:efe832ed8d456afa7a7184cc49c6e860d6deaaec}}, which was a recent breakthrough at the t... | m | 24a7d7f31c9cd0b29e03f1a17a2962ac |
Spin dynamics {{cite:3c02bc7f1edc29dae8ad4585ade3d25f889fa23c}} are also regularly used in network clustering. The first step is to define a spin model on the network, consisting of
a set of spin variables {{formula:9df86946-9983-4b5f-9c80-a46f6692fdd8}} , assigned to the vertices and a Hamiltonian {{formula:d01a6d04-0... | m | bb76b847446496190f955164fd00a63a |
Furthermore, in practice, the phase-shifting of each PS can only take finite discrete value from the discrete phase-shifting set {{cite:9e501e56cc169722289ed477d3cabeb2440301d4}}, {{cite:c671941b61e3bbacdf94b97bf3df55a60f7bfec5}}. Let {{formula:ba5f5703-192b-4f92-95cf-d18267c7fda4}} denote the number of bits to repres... | d | 327144bca50c84b670ee0af9d4c2c369 |
Many deep learning methods have been proposed to generate time series, including Recurrent Neural Network (RNN) based methods {{cite:7d1025ee22be1458bc3c6445f9103bd1287f1d12}}, {{cite:887e9403060ddce12018470bba919fb2f2d6c1e7}} and block-style methods {{cite:5e45808bbf420d68597c7538d6b51293cd597158}}, {{cite:9c87481e407... | m | 458c4a193a3fd8c83ef047978d9b51ea |
During outburst, the RG is engulfed by the ejecta and the near-side is exposed to the radiation from the hot WD. The immediate effect is to increase the distance at which dust condensation is possible. {{cite:fd70b2492aa147451f8fe4049ecefef208547e15}} showed that the sublimation radius reaches a maximum on about day 70... | d | 08856ee5f1af102ab8542a0063f66c33 |
On the side of adiabatic quantum computing {{cite:fa5a012c7dd93418ec0443c46f8aca4154d8c597}}, despite the fact that it has been commercially available for more than a decade in the form of quantum annealers (QA) {{cite:8ff3d6b50bd7442f9d0d3495d8ce14b546bc0d3e}}, this approach has only recently been used for LGT calcula... | i | c72b0585f2c2a85cdc3e41e518db995e |
Radford et al., 2015 has shown a way to build high quality representations by training a GAN model and reusing parts of the generator and discriminator networks as feature extractors for other supervised tasks {{cite:c30331ee3c6c125ace295bd7819e7af5d584f76a}}. This can be potentially a promising future direction for th... | d | c40f02dc6ea4b553eabec949049ed77d |
The IRL model framework is illustrated in Fig. REF . Model training can be conceptualized as a Policy Generator (G) and a Discriminator (D) locked in an adversarial process {{cite:765e68a281e9adf415a90a60e7860a5be1d0f81e}}. The Generator generates fake eye movements (Actions) with the goal of fooling the Discriminator ... | m | 6402f7a2061eda391d06e77a26f123b7 |
From Lemma REF , we know if we get an {{formula:4cb007f2-cefb-4158-958c-4125e6b68d0f}} -biased set of size {{formula:6c319d83-c18f-4e5d-86a2-38f424e7c93b}} , we can use it to recover {{formula:1021b376-77f0-47b4-9b14-cc57d4d32922}} , by making {{formula:c2b1c0c9-3406-4ad9-b4a6-14530a8446e5}} queries. Thus, the key tec... | m | 5df17f6ccf5b9d41212f0f2658a1c10f |
Let {{formula:7db5527a-14ce-4450-bb33-114b566059bf}} be the vector bundle on {{formula:f1148b8b-938c-4c95-8700-829b72fa2697}} arising from the {{formula:90767d07-5337-46de-a405-c078ca9d35e2}} th step
of the Hodge filtration for the {{formula:dd3afc52-a7ba-4b76-9468-74cba11d46be}} th relative cohomology of {{formula:2... | r | 1b31d2a0373d0be729a363e1b8634c39 |
The first part of our experiments (sec:Experiments-Quantitative) comprises a quantitative evaluation of the accuracy and performance of our optimized implementations on the KITTI 2015 stereo benchmark {{cite:8b108117aa9ef8297cdc90c00279e7799090bef4}}, as well as the Middlebury 2014 stereo benchmark {{cite:f9832be4cfb6b... | r | a2ef96cc519864e3a5015d25f711d1a6 |
Limitation. 1) Currently, our method does not applicable to multiple input voices. 2) Since we do not focus on fast inference, the high-resolution rendering process cannot meet the real-time requirement, but our model can benefit from some concurrent NeRF acceleration works {{cite:54e4c4af9cf5adb92b414007fcffbc126b4540... | d | 1c38d4295613bff4ca7600ecc5835a47 |
There are millions of photos of natural landscapes on the Internet, capturing
breathtaking scenery across the world.
Recent advances in vision and graphics have led to the ability to turn such images into compelling 3D photos {{cite:449dc0d3d8238d40adb58aba785c27e1f6fbc7ac}}, {{cite:69f7e4adbc1b1614312b0db86650a4b7bd7f... | i | f13aed5d6f077e649f154dbcff95b1f4 |
The Tikhonov method {{cite:176d7cc00ea1bb184a3c29157f0f04e9da841c73}} is to solve the above question takes the form
{{formula:e3766b73-33a5-4bd4-b56e-fc5e55db0b55}}
| m | 19b6b0f8eb5e7a7ab21c909b76aa56f2 |
Actually, our counterexample shows more, namely that the following substitute of (REF ) for “long-range" potentials (i.e. {{formula:4217ea53-a871-4fe1-a413-05b070020b6d}} ), due to Frank {{cite:0fe269e938de11212bf6592ceb67b4fd630e70b0}},
{{formula:85049b10-2cd4-41a2-97a2-0c39f52d69da}}
| i | 60da32b666d24faad98b0d413f84f33f |
DBSCAN is an effective algorithm to discover clusters that present unusual shapes even in large spatial databases{{cite:c01b1387ed947963e265dfb6062f998147c9beca}}.
| m | c118bdd291ce262e54d78dc2c6ae832f |
The recent experimental discoveries of superconductivity in rhombohedral, or ABC-stacked, trilayer graphene{{cite:570c35396383065ecde6b9ce23eeaa1b10016913}} and in twisted graphene bilayer systems {{cite:925557e0ae3cfb13ebbc665b81ea2c88649e31c6}} have received a lot of attention. A vast range of theoretical proposals h... | i | 5a83482e76ba4d5188bb78e08aee9f5b |
up to subsequence, {{formula:24ce8d56-7365-4222-ba7a-88a2586993cf}} converges to some {{formula:eee07300-a8a3-4a5b-b54c-394179c4ae6c}} with {{formula:79f156d4-f35f-4455-9161-fe4649078ec1}} . From {{cite:3807fa58e8d777edcc7d3daf720563040d62245d}} we conclude that {{formula:73125aca-9eb6-4753-9160-50b4f233b91e}} is a ... | r | 019a98fae9d43d42f0c3ab16e701ff71 |
where {{formula:5ff33d99-f7ac-4da6-90ee-9ebcddf1bf1d}} is the minimizer of {{formula:de7977e3-38b8-47be-90b5-9397c3b88a32}} obtained in
Theorem REF . The function {{formula:e48a3610-f4eb-49eb-97ad-15285fc65d4d}} is named as the
regularized solution to (REF ). Let {{formula:c705cffe-1197-47ec-a3e9-4f2d613bcb0a}} be ... | m | ea7a204b48535715641865e4a724d8dc |
We evaluate our method both qualitatively and quantitatively on video pairs of walking animals from the YouTube-VOS dataset {{cite:53e60f0b45cdd2a93d103366fed9a42aec68fec6}}. All pairs are challenging as they contain different shapes such as a cat/fox or a deer/horse. To demonstrate versatility, we also present results... | r | 40c7f9bbd67442e15d3080cf9c35ce82 |
where the sub-kernel {{formula:d188f2e1-17d1-461b-a463-a269842a0ae6}} is given by (REF ) with {{formula:e5bf6ae3-fd27-4243-a27c-15812310a85e}} and {{formula:aa3f66cb-fad4-44a7-ba20-5c3d03397923}} . Thus, problem (REF ) is transformed into the problem of determining {{formula:ee60d8d1-faa1-432f-b4d8-6f5f22fd5399}} a... | m | 817c96a9852e81a84ff07d7661ffd125 |
This theorem enables the estimation of the gradient of the value function analytically, whose evaluation in the context of sample complexity scales only logarithmic in the parameters {{formula:b16bdfd0-197c-4170-b0e7-eeadad274bcb}} of the policy {{cite:1ff7a82ffd1f176f0eda1f4f95bfec51038b4352}}.
| m | 792fc47f0a6534df0762e91ae1076a85 |
A novel method for cell detection and counting, called ACDC, was proposed, by exploiting a fully automatic pipeline based on the watershed transform {{cite:2546961f85cb10b8679d9d3fcec894d05e6d2fde}}, {{cite:0376060002e12440d31dff88a4b4972c44574849}} and morphological filtering operations {{cite:dd685c6fba4802c925a02422... | d | 683c0e15593553920f042602d8568f6d |
While there are many definitions of robustness, the focus of our work is on distributionally robustness, that is, being robust to the distribution shift of the local data distributions. We consider a distributionally robust learning perspective by seeking the best solution for the worst-case distribution. Another key f... | i | e412750fc1f0d169a5d9ed1a8db5b02f |
In recent years, several studies{{cite:50ef947c1b78ab2a1c0c835025a7274dffb2c5be}}, {{cite:8f74bdb719dfa79269b5d8b76de383ccdd6804a1}}, {{cite:24eecb3cb9a6f5be051335640a01710af8b6f82f}}, {{cite:3a732765b2e10e1fa009fd907e2e41b783bb1033}}, {{cite:713680dd26582615fe8a53d4a57203394a648cc0}} has been proving the efficiency of... | i | 58128bffbf726d58b5535db7b8ba6c67 |
Theorem 1.1 {{cite:5ef56e749bfdb2d489f19fea6ac7252862e9d3a5}}
Let {{formula:196465b5-50cf-4ac9-bd25-65dc3de92602}} and {{formula:23aa2c97-c57e-44c1-ad67-1ef24791a7ed}} be separable Hilbert spaces and let {{formula:272b8214-5f82-437f-ac43-5d433387c744}} , {{formula:548cb672-2212-4629-8070-447efd35cf63}} . The followin... | i | 5e08e1277ca50b7f807b2cabd4eb3f93 |
This subsection highlights the results attained by measuring three different metrics to judge noise reduction and the quality of the reconstructed low dose CT images. We use the following metrics for the evaluation - Peak Signal to Noise Ratio (PSNR), Structural Similarity (SSIM), and Root Mean Square Error (RMSE). PSN... | r | 90c191848551a1f796d6f58fef787e0b |
Human intelligence is often understood as a process of turning experience into new behavior, knowledge, and skills. Humans achieve this by constructing small models of the world in their brains to explain the sensory experience and use them to infer new consequences. These models can be understood as turning experience... | i | a6c7afb0a4f471011156811702f76cda |
The idea of variational inference is to optimize the parameters of a variational posterior in order to come as close as possible to the true posterior. Following {{cite:9300e75b0a997de748141a8de7d4e1a7d85529a4}} we choose a Gaussian variational posterior,
{{formula:d4a4f95d-d875-4306-b8b2-b962541f614b}}
| m | 87f82d4d8edf09bfa5ef6039b00694ca |
We primarily consider methods that utilize GNN models which adopted messaging passing operation as their main backbones. GCN {{cite:c2b2eeb70d28ea17a03f2347947179b6d089141e}} and GraphSage {{cite:cf392d566b0af58c3805b446783335a265557c23}} are methods where convolutions are strictly based on first order neighbour aggreg... | m | b5b67e9a93e73902dfe05b5d2dd76151 |
In the last few decades, RMF formalism emerges as one of the most important and prominent theories capable of interpreting the finite nuclei results of heavy-ion collision experiments and the data obtained from the astrophysical observations adequately. A lot of RMF parameter sets has been developed in the last few yea... | r | 8979962968ca59eb026508400d7ef94b |
The protocol setting is as follows. The employed DQN {{formula:0cd70c5b-ffab-4e16-a4a8-768e99e6f04a}} consists of two hidden layers, six residual blocks, and {{formula:5a1574be-2fbe-4174-9693-89c3ae255045}} output neurons, where {{formula:3338f1db-f497-431d-8465-c57acd9093a2}} for Clifford+T group, HRC gates, and Fi... | r | fe5db9d63c75cc3be180f58d0d232afc |
However, without principled guidance, the underlying mechanism of transfer learning is not very well understood. First, when we are pre-training deep models on the upstream dataset, we lack good metrics for measuring the quality of the learned model or representation in terms of transferability. In the past, people ten... | i | 43228449212499695ce04814678629f9 |
Based on this claim, Cut&Paste {{cite:b2426ba7120ddce447b5c8844c2c2a50ee272027}} developed an SSL model in which the proxy task is adapted to detect irregular patterns. They created a data augmentation strategy in which a patch in an image is copied to another location after being randomly modified. This data-driven st... | m | 31795b361a4b668040daa888b0fefae8 |
In order to determine the nuclear modification factor {{formula:55b30261-2ef8-4f82-a0f7-78d181d0c38b}} , the interpolated {{formula:b7d5189f-8990-4360-922a-ba84748d8da0}} -differential pp cross section is scaled by the average nuclear overlap function {{formula:fb9ec8fe-0ec2-4028-9e1a-f23e6b88df92}} .
The resulting nuc... | r | bbd5d529ce4241a5f1eac4733f5a73e3 |
In the Merton jump diffusion model (MJD), see {{cite:6165cd54d57fe1668bdc49543ff0da2e97ad34e7}},
which is a special case of the Bates model, see {{cite:2c326e0bc6998556d4b3f74c90512a6ad86af68e}},
the stock price is modeled by a jump-diffusion processThe number of jumps are modeled by a Poisson process with intensity
{{... | m | ced57388893d51211adee6d73e8b9e05 |
In this paper, we introduce anomaly clustering, a problem of grouping images into different types of anomalous patterns.
While there has been significant progress in image clustering research {{cite:5ad2db06160b71f88b58e853c2e5ee37e66609ec}}, {{cite:d73251348631ce73f77e639c87375fd425bc6df6}}, we argue that the problem ... | i | c827e2f85a71c48ceda289cb1b0572c7 |
Nevertheless, in recent times, a lot of research has been done to understand
the Hamilton–Jacobi equation from a more general geometric approach,
and some geometric descriptions to the theory were done in
{{cite:06cba4f87ec0e73f16e2bed2f4f5b1364d427f70}}, {{cite:77f64cc303faf84e1cb8b30a2a74b984a0a8a55b}}, {{cite:4d39a1... | i | 0299d2e4cdba54436749342ae8c8041c |
In our simulations, magnetisation dynamics is governed by the Landau-Lifshitz-Gilbert (LLG) equation {{cite:65c63ee7e3ff2e1a375b6f4bfbb022dc9d73361e}}, {{cite:b84ca280a315a5c83267ebf4ffeb5b081a4c3bbe}}
{{formula:15e4ea1d-a1df-4ec8-9411-59ce099b6a5f}}
| m | 9ce0964a327de32f1a0a131b362b0838 |
Based on the complexity of the SQL, the examples in Spider are classified into four types: Easy, Medium, Hard, Extra Hard.
Here, we provide a breakdown analysis on the Spider test set, as shown in Table REF .
The BERT results are adopted from {{cite:c3287c6b9fd90b1ceeed1bcb445c6fb4ed783011}}, which is the state-of-the-... | r | adf480640acaac8ba58ab9afb2e40aa0 |
Strongly robust toric ideals appear
in geometry,
as defining ideals of non-pyramidal self-dual projective toric varieties, see {{cite:b73686519822d67b8965e22175881d7bd2c6f251}}, {{cite:dc8fb0879db6f02e53475490a736af5da8670b58}}; in combinatorial commutative algebra,
as toric ideals of Lawrence matrices helping in the c... | i | e9f3ae7c4232a8fd4dcef451cf8a10e0 |
First, we perform a linear evaluation by fine-tuning a fully connected classifier for 100 epochs on top of the frozen 2048-D feature vector from the ResNet-50 backbone. Table REF reports the results on Imagenet1K after 200 epochs of pre-training. With different batch sizes of 256, 1024 and {{formula:54bce379-db2f-4945... | r | ff1f90cb6c2f22a025876a47d95f2740 |
Dynamic models are based on two levels of specification: time-indexed observables {{formula:4f335289-36b4-40c4-af30-c9c87b65a219}} , {{formula:84da7fcf-2afa-492e-8dc6-11444392e722}} are assumed conditionally independent, given latent states {{formula:53d7e17f-65f7-416d-847a-ab4d3df04a93}} , governed by stochastic evol... | i | da1ccf2e1f6a3c19ef19a81f60ec720f |
It is noted that in this discussion only models employing leptons as primary particles are considered. The stellar
wind and the circumstellar disc provide reservoirs of target material with which relativistic protons could also
interact to produce gamma rays through {{formula:026c2ba6-ec2e-4cdf-8ef8-dc14c9c08182}} dec... | d | 0015eb82506b550325761f2b11008298 |
In another context, it is often not straightforward to choose the most adapted statistical divergence for a given application. Indeed, there are many types of statistical divergences that are indexed by a scalar parameter, for some value of which the KL divergence is recovered {{cite:3663f250f219aa326e1990b1c2bd656c908... | d | ec00173c45b936c87372aae1e6eb9094 |
The fast modes most likely originate as sound waves from the photosphere that mode-convert into fast modes at the equipartition
layer. However the fast mode wave amplitudes quoted above at the equipartition layer are significantly larger than those expected to derive
from waves associated with photospheric convection. ... | d | 917a4f30c29e381a41e169ec16eacd0c |
All models are trained with the Nadam optimizer {{cite:74c86ab19af713e3da41e6c7d7f399fbc04ceb5f}} on the training set and the model with the smallest loss {{formula:1da90bb7-c46c-4c8f-b77a-f9e789d73638}} on the validation set is chosen for evaluation. 5 training runs are performed per model with different random seeds... | r | 181d839978bd57fb5111ca7a3ae88ae0 |
Different auxiliary property-based tasks can also be integrated into a hybrid method. Hu et al. {{cite:ae57fe20e924793a20f4aa25eb7622527c9f2209}} present to pre-train GNNs with multiple tasks simultaneously to capture transferable generic graph structures, including Denoising Link Reconstruction, Centrality Score Ranki... | m | 4d8a93207a40d05c9b76e8f58857da83 |
Statistical physics offers a way to pass from a microscopic description of systems to a macroscopic one. Its numerical realization is known as “molecular dynamics”, which is a simulation method used and developed over the past 70 years; see {{cite:21fc44ac9282dbe2e4f2efef344690691e24645e}} for a historical perspective,... | i | 6803d0b336cd70e1ec0dd0642c6a7822 |
Label sparsity does affect all relation categories, apart from one-to-one category.
We further investigate the changes in system ranking by different relation categories when the label completeness is improved.
The relations in KG can be commonly categorized as one-to-one (1-1), one-to-many (1-N), many-to-one (N-1), an... | r | a19851a774f7410f46221ab3c1376982 |
This second-kind superconducting phase transition {{cite:2efafe4dd1c4a006a6014907a48d7ebeb0b30453}} is known from low temperature solid state physics {{cite:e77db8b4ff0471c8d9691a66e1965de35091baf9}}, {{cite:c6e5e8e8662dad9c3fd08581db14c8fae1b68f29}}, evolving Meissner diamagnetism based on electron pairing and condens... | i | 8c631d1f87ce70afa8ec5253ff113eb2 |
We find that the AHE in EuIn{{formula:f3177445-fc29-4775-9e67-2fe5287f7a56}} As{{formula:ce13b034-edcd-412b-a8bf-7333ad4af9ab}} fulfills these features of the intrinsic AHE.
We plot the anomalous Hall conductivity {{formula:ce90f02d-ee81-42c1-a897-3ac2b65c9a6f}} against the longitudinal conductivity {{formula:85e5a6d... | r | ef97738922eabaf2abd99af0cb730924 |
The formalism will be exploited in a companion paper to investigate the Hayden-Preskill scenario of throwing a diary into a black hole and asking when the information is returned in the radiation {{cite:75a709b93e132102eb1dd6231e4911fce1315ad5}} and the Harlow-Hayden process of distilling the purifier of a late portion... | d | 372020834d464474e8b196df51eff1d5 |
On the other hand, the idea of using the Morse index of a solution for a semilinear elliptic equation was first explored by Bahri and Lions {{cite:5055bb1087bf67e40a98dfdd7cf838df2b6a80c3}} to get further proved that when {{formula:dc6a5afa-f3f0-4d01-8ae1-98c17b996b56}} , no sign-changing solution exists for (REF ). To... | i | fed70707c048d72bfff6355ec2d5dfe1 |
The vector space {{formula:ad9c81a4-c257-4643-9b28-7c09f9be99bf}} was explicitly calculated by
Peterson {{cite:5f60a6e1fb0b28e301f18f703fd6271458edb23a}} for {{formula:1726cb63-140f-4f51-8123-c9197336dac2}} by Kameko {{cite:ede054b207f12e5caf75b0d16e9768c6129cbb4c}} for {{formula:95865e1b-869a-48bc-aa70-2c8f40ab0ea8}... | i | c79b6518e8dc3fe7d3431e8f0d4ccab6 |
where {{formula:8166179d-ff12-4cc2-a979-90ec42530ea7}} , {{formula:8b18251b-e48c-480e-8394-63a18f6d1d19}} for {{formula:7fda118e-0dd9-4704-960a-96a42ba3b1ce}} and {{formula:55370c8d-e85e-47d6-a9ea-ccc9cbb5481f}} .
We remark that {{formula:451b158b-aa61-445a-96b9-dc8892ac936a}} denotes {{formula:6c4b8221-42a8-4b30-91... | r | b1affcb50a441a44eb1f6e10fb791441 |
Overall, HaloAE performed competitively on the MVTec dataset {{cite:233a75e69c85603ec9a5fe8dcd45178ced5ab283}}, with an average score of 91.4% for image-level detection and 91.2% for pixel-wise segmentation. The performances of our hybrid model between CNN and local Transformer suggests the importance of integrating gl... | d | 416f8f074819a815936a9b040a9e8018 |
For MedMNIST3D, we implement ResNet-18 / ResNet-50{{cite:679155bf94eb6da7e246dc30d6c24375ea94cbe6}} with 2.5D / 3D / ACS {{cite:e3e74d06916267ba6cc57495b3640bc01f823fac}} convolutions with a simple early-stopping strategy on validation set as baseline methods, using the one-line 2D neural network converters provided in... | m | 4974f866fe00da74ce5b856e15dca82f |
Furthermore,
{{formula:f3f17887-5704-43e6-8cd2-275c375b3537}} serves as a more direct signal to characterize locally how
vortical the quark gluon plasma is and we may extract the magnitude
of local fluid vorticity {{formula:9ec469ca-8691-4dbf-829f-75838fc4b835}} .
In current study, we find that {{formula:b5f8b379-6c80... | d | 09b77a3da919c5fa83b6cb8ce19acb5e |
Other time-sensitive protocols.
While the attacks we demonstrated were specifically targeting Lightning Network, we believe that a wide variety of Bitcoin second layer protocols {{cite:24d4fd00c68bfcdc5ee19b6f94094f3253ea6053}}, {{cite:f28f2eefafdcf502772adf9d404436d002a5f9d9}}, {{cite:11c4dd874aafbfe959d9b741cae37b51e... | d | 99e386936fce8159da17607a7f5c846b |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.