text stringlengths 54 548k | label stringclasses 4
values | id_ stringlengths 32 32 |
|---|---|---|
More generally speaking, inverse problems consist of finding the unknown characteristics of a structural system from some of the outputs, or measurements of that system. Most notably, this includes the above mentioned inverse conductivity problem in geophysics {{cite:01e35cd44f4e5381b503fbda63ea937b594c1b19}} and engin... | m | 8f29bd60989b267e5fc036109c56430b |
However, in the presence of low-dimensional intrinsic structure and well-behaved noise that is symmetric, sub-Gaussian and independent of the {{formula:c393b32d-e5b3-463c-a7ed-5069aeb85188}} and {{formula:d1d11dad-c18d-47b9-a011-8496250c8a4b}} , RERM
with optimal, large enough, tuning parameter achieves a much smaller... | d | 91fddd97c69f91e3fee69c6e692ad672 |
In Figure REF we compare the evolution of the axial ratios {{formula:1b0bfb64-766b-4bd4-96cb-784eea6d67bf}} and {{formula:9372b075-f11d-4af3-bf56-69e2668538d7}} and the anisotropy parameters for {{formula:6a93d0ef-5b9f-4cc2-b7fd-5f618fdf18bb}} models with {{formula:8299e451-0f83-4520-8473-1929b323af5d}} and 1.8 fo... | d | e8c083c6a949f5915c1099ccee229c42 |
Our method has introduced a significant step towards creating truly realistic 3D talking heads, as it has been shown by our extensive objective and subjective evaluation against other SoTA methods. It is important to note that our method even outpeforms DAD, which was trained with 3D annotated data on a large-scale dat... | d | a2174b80dd79538e6634ba1b82d84050 |
with {{formula:e4206ee1-954f-474e-92ef-09e85bd5cad4}} . {{cite:1865069f80df0fb444997c9f3bb79210614a081b}} extended this view of eigenvectors to derive sparse loadings by implementing a regularization term. The results are based on Theorem 3 of {{cite:1865069f80df0fb444997c9f3bb79210614a081b}} which we recap for comple... | r | e03f4da9bdd1ec4586a152e736ba07d2 |
In the energy bins below 8 EeV no significant measurement has been obtained for the equatorial dipole amplitude {{formula:a127a37d-9b08-40d3-8c20-a4035092bff0}} , and the 99% CL upper bounds on the amplitude are also displayed in the plot for those bins. For both magnetic field models considered the predicted amplitude... | d | fb728b3bb5e8a569a413492d8c1ab5aa |
And a CFT module has 8 duplicate Transformer blocks, as shown in Fig. REF .
In addition, since the dimension of the input sentences {{formula:610f3247-9079-41fa-bad8-6baca3972bd0}} is {{formula:ef463880-86b6-4635-aef0-05516ace1342}} , the actual expression of {{formula:69d31218-a94a-4984-b68c-6191d492536c}} in the a... | m | 63ea3ffdf5db655bff82b0e074b8f7de |
where {{formula:82f8eaf8-dfd8-4736-a267-6dfc811e3291}} is the principal branch of the Lambert {{formula:a77ad778-e176-456a-981d-20e6f7168e39}} function (or product logarithm). A key property of {{formula:008fb2cd-c7df-4c0c-b1e3-50dcf98d8823}} is that {{formula:8f099d40-637c-4b91-bfed-2f2762ad4e78}} for all {{formul... | r | 03267d0bb193beb020abb90d5e178731 |
Spatial Graph-Convolution The purely spatial convolution is computed using graph convolutions (GCNs) as defined in {{cite:b06ee98d55dd5100f24f594126bda4453f96c809}}.
Instead of using the full skeletal graph, we again utilize the kinematic tree that was described above to convert the undirected skeletal graph into ... | m | 556f0cfa2d26607d4836096e49092ae9 |
Because of the inequality {{formula:baf6167a-a706-4e6f-bc21-37f43067067d}} , the second integral can't diverge to {{formula:6a6c38f0-a0ac-4b1f-b00a-1e649aad954c}} and therefore for any measure in the Szegő class we have {{formula:bd7453b6-5ea5-4148-bccb-450e10fe7691}} . Furthermore, in this case there exists an outer ... | r | f23683feab2d7754700ed70e4558b036 |
Finally, the three-dimensional Minkowski Functionals contain more information than their two-dimensional counterparts, and a complete analysis of the three-dimensional field will be forthcoming. In this work, and throughout a series of papers {{cite:77649118239cd63a5bd334f223e43c1ea9fb0ee6}}, {{cite:d451a0d64857b87f2ba... | d | 2a91b47434741f5e3454c229dda36097 |
We provide additional results of conditional image-to-image translation (Conditional I2I) {{cite:1b14dbd26bef44974299934180efbe5e89799e3a}} and style-guided synthesis {{cite:30ed39489998e8e810f0c9fef01cb203eaa1ca8f}} in Fig. REF , column 9 and 10. To train the model of {{cite:1b14dbd26bef44974299934180efbe5e89799e3a}},... | r | 6f86b47cfc2e0e32dec0746e2cfa9ae8 |
As mentioned in the introduction, the anomalous decrease of the bandwidth near the magic twist angle promotes the importance of the next-to-leading order terms in setting the anisotropies, thus selecting from the nearly degenerate manifold of correlated states that are obtained if only the leading order terms are kept.... | d | a1f2ec41b55327505a08e5093cc3516d |
Instead of comparing the performance of the ERM against the optimal risk attainable in the class, one may wish to compare against the risk of the optimal 1-step-ahead forecast.
For loss functions in the Bregman class the optimal 1-step-ahead forecast is the conditional mean (assuming it exists) {{cite:ad371033712b723cc... | d | 4fb246cdec4a613b7dfa0ed3b99beda4 |
Meanwhile, several training techniques {{cite:e3d1b41ebcc85973634a1a4f03c915d5a7c3754c}}, {{cite:41bb5841a62567148157058c7c66d8403747ce46}} have been proposed to derive a more robust embedding model against the perturbations. On the one hand, data augmentation methods transform a single image {{cite:1445341a750b91b9033... | i | 0c95e53b27aeb93554e0043534fe47a1 |
The majority of span-based models {{cite:1ef2004f8fa338dc7cf6a9af91f7ebac859004fa}}, {{cite:a7ff7f19048c670a3a98580b3e883a43d200057d}}, {{cite:0d91205a6bf018bb768d4e19be14a36c90049614}} use Pre-trained Language Models (PLMs) as their encoders directly, which relies on the encoding ability of PLMs heavily, resulting in ... | i | 4b6abcb7b23ab6da8858fcf16a2792fb |
We also analysed participant feedback in task 1 by means of inductive thematic analysis based on the framework by Braun and Clarke {{cite:d36bf42f3a1969d5ba7e8eca1a631235e64c5144}}. The goal was to identify common themes regarding positive and negative factors considering the three videos created for task 1 of the user... | m | 870bfec43b2bbf702bb30fa8470bd91c |
As discussed in Section REF it is very likely that there are more general supersymmetric black holes with an additional electric charge corresponding to the {{formula:fc54ee1b-60ed-4e59-ae6b-2775a64fb804}} flavor symmetry of the class {{formula:0d86ebc3-76a5-4f09-b43c-e30fad4b035f}} SCFTs. It will be very interesti... | d | e5da26b22f1bfafd308d1f32f298dd28 |
Such an approach is convenient in some circumstances when it provides physical insights on quantum field theory {{cite:eeb610177504e33874f2da3643f617144c6d462a}}, {{cite:284ddc44cc096ab36c16713117198e8cb756aca4}}, {{cite:72ee64c4fe91fe7c68c8a262584677b5f69ffa71}}. For example,
an appropriate version of OPT is given by ... | d | cc6d76a45dd953e487483667a7e3a758 |
Our approach is fully differentiable and can therefore be trained end-to-end using multi-view images.
Our experiments show that when trained on large amounts of data, our method can render high-resolution photo-realistic novel views for unseen scenes that contain complex geometry and materials, and our quantitative eva... | i | 6732164ca8024f1a7fab82412947b693 |
Generalization. As explained in Sec. ,
different types of media data (e.g., human face vs. vehicle photos) are
generally projected toward distinct manifold
spaces {{cite:5d35a2c706e23e8ec335675661435987eac7b08b}}, {{cite:a6eadf43ba7ee4dd9540efb4fb39ea976025136d}}, {{cite:a5583311c243970c57ab4d9aca999a513098d403}}.
Henc... | r | c66e309e0d4cee38cfd8b4d535010103 |
In {{cite:4e4cd36db67226b3ce9e69e45b4c935aa9264b17}}, {{cite:a98df987eeed9aa48b509a8839c29c5749c2e449}} a sequence of probability distributions
{{formula:4e063e1f-ead7-41bf-bcf7-2130a5317c3a}} is said to be chaotic if, calling
{{formula:8a8ab092-f539-4fa3-8bd3-7b4db02d6071}}
| r | 6a23ed0015c8ffea9e8f634ef31a3c9d |
Due to the evident domain gap for vision-based detection systems, some works consider the auditory domain instead of the visual domain for object detection. The auditory domain does not exhibit a domain gap between different lighting conditions, which makes this modality robust to such environmental conditions. The tas... | i | 9ffd23f93096533586d940989b842cf2 |
The LTS schemes in {{cite:7b0e6d67a45d294283ce007caf38c6838d834810}} are based on explicit methods of second and third order accuracy, normally referred to as strong stability preserving Runge-Kutta (SSPRK) methods {{cite:99a1778b0004604468c770b89df0df02414c7fe0}}, {{cite:fe16f4ae4d2cf430a68fa7d1c8f0d6c5130e0ef3}}.
The... | m | 8c2f8eecef69bc218d275bcb66d6057b |
In this work, we have built upon the techniques used in {{cite:1c9c1692b56d6add8c2b72c4376b10655079360e}} and presented a graphical framework that constructs complex codes from tensor networks, which may be considered as a natural generalization of code concatenation. While holographic codes were among its first concre... | d | 84183f1d46d88a356d098e6ae18335e6 |
with {{formula:4798771b-d7b5-4771-9f98-18846097e849}} satisfying the so-called second Heavenly equation {{cite:3909bebc7807a9469b2af88f3d2d63114b12fad8}}
{{formula:e29d8491-6e85-4860-ac06-1ef80ff8a7b1}}
| d | 14785b628398f6f1ac0a098ff20d3879 |
One interesting question is to construct the characters of these
representations and hence forming the partition function. One could
further compare with the partition function of the Euclidean black
hole {{cite:718ebade74d0f356a896a22bf1958b6a5e9f3555}} and trace the winding modes to the Lorentzian
geometry. In this r... | d | 01f166f9e7d92fc893d14362649b857d |
Further suppose that, in the graph on the right, the causal influence of V{{formula:2377dc8e-3966-4e8a-8848-b5e3f375e275}} on V{{formula:d3354b65-4024-4760-9c55-e85728412c86}} is precisely canceled by the causal influence of V{{formula:edcf9745-da8f-49bd-b7ff-603a97474a51}} on V{{formula:d29709d1-8bd4-45ec-a16e-b62f... | m | 448a6e04468d2c07d75f84df9e5f482b |
An XAI method {{formula:884abeba-9d03-45cb-b759-caf6f0a874cd}} takes a model {{formula:bbdcf6e6-4d55-44e3-a514-f09ff2d9b057}} and a molecular graph {{formula:08537901-bc41-4b88-9dd6-4cdbe0d13028}} as inputs to generate an attribution score {{formula:7dce8fbd-6c55-433c-95de-1259915363e0}} where {{formula:67ac0658-f7... | m | fd60bdc9f5ed6116922a6b3756bcbe48 |
Under {{formula:b1c1a377-0cf2-4402-b68c-c83ad7d7b754}} , the test statistic
{{formula:fdf9de05-44a0-4bd0-ad6c-765293afc40c}} has an {{formula:b5b632f7-e431-4dc2-b0d3-d18443d42a6b}} distribution with
{{formula:3a3f0b44-98c1-46a4-9f6f-8a97173e7170}} degrees of freedom,
where {{formula:8b3b5a0a-23d4-4727-b8e2-d8b75854b... | m | 6e92bede5e5d789821172865b4258daa |
Additionally, we report UCE values from a DenseNet ensemble for comparison.
In contrast to what is reported by {{cite:95fad12b334727ed0b59a803c0d12e9167d9bc2a}}, the deep ensemble tends to be calibrated worse.
Only on BoneAge, the ensemble is better calibrated prior to recalibration of the other methods.
After recalibr... | r | f1dd65d7e0ff5fe2f43a6965ec40362f |
Furthermore, we compare our result to the condition in {{cite:831c5ca6707cb11d782da218962eb974dea26240}} that involves a bound by the algebraic connectivityThe second smallest eigenvalue of a Laplacian. {{formula:cd5a2dc8-2275-4cdc-8109-d323152c1aba}} .
We compute {{formula:65f4851d-a034-4298-81fc-c351b9129ed2}} where... | r | 4af639fd33e97921f0f904ed24083c21 |
is precisely the sum of Dirac measures located at the eigenvalues
{{formula:86147aef-6a03-43ee-99b4-0065608e2dff}} of the operator {{formula:a18a43c4-ad16-42a1-8462-8dc027f95015}} .
Similar results were obtained for various classes of differential and pseudo-differential operators in {{cite:668b815079aa86f08392407b80b... | i | b8784fc64617ef505df9036f0ba54395 |
The present data-driven approach was demonstrated with a set of local sensor measurements.
As this method performs spatial field reconstruction at each time, changes in the number of sensors or motion of the sensors are easily accommodated.
Having this flexibility allows for extensions to incorporate other types of mea... | d | cd296d3463df2f6091a9597eb3e56027 |
There are infinitely many sets of context-free grammatical rules, and thus it is not appropriate to assume the uniform
prior over all the possible PCFGs (improperness).
This study adopts the hierarchical Dirichlet process (HDP) for the prior over PCFGs {{cite:130aa7d01a13eaffa7ab9ac37e9851ff42e9791c}}, {{cite:6c93687bd... | m | 16eb52f10a9d3a5cdbbd239df85c2f9e |
Condition REF is usually imposed in the analyses of high dimensional linear regression models {{cite:7a0cdd841d9e7ce4491be7608e0e8ecbe9752da8}}. This condition implies that the sample covariance matrix {{formula:dcaa41a3-2150-45aa-bc90-55c614d087d8}} satisfies a so-called restricted value condition over a cone set, w... | r | 9b59fb82fb7d17d01bb5a5ceec664d84 |
In section and the Appendix, we outlined and revised the necessary formulas for the AFCDM which allowed us to construct exact and parametric 8D solutions with a fixed energy parameter in the conventional co-fiber space of the phase space modeled as cotangent Lorentz bundle nonholonomically deformed by a nonassociativ... | d | f404147218f667a0957f79bb2d799b12 |
Base: Each client uses local data to train local models.
FedAvg {{cite:f36a9e06fca1c4f7a1cd48540131645be8a1a2b1}}: The cloud aggregate all client models without any particular operations for non-iid data.
FedBN {{cite:eba658855cb8ffa4e231ed7752072c36f9439b06}}: Each client preserves the local batch normalization.
F... | m | 58ed5c0c7059444d79dc0dccfc778c88 |
Question types. In tab:mainresult, we demonstrates accuracy on six types of questions based on their prefixes. Most models tend to perform better on the “Is” and “Can” questions while delivering worse results on “What” questions, likely due to a smaller number of answer candidates – most questions with binary answers s... | r | 8fe12354e52f7ee3ee704ded1bc31409 |
M2 SMOTE TomekLinks {{cite:f7527b03d9bfd2d58b7a7c8186fdebfefda7f158}} 13 Binary C4.5 AUC
| m | 23cee281446eb94c343f5799a94debec |
Note that without the constraint of being in an LOCC protocol, a
measurement on a {{formula:67461186-8e37-4a30-a83e-228040448437}} -dimensional system can be compressed to
{{formula:09a7f14d-c2c4-49b9-b053-c76568c29e05}} outcomes.
This bound {{formula:481f47e4-d60a-474e-9c4c-b52e0acca9eb}} shows that to optimize stat... | m | 4d2edb0feefcc45bbe66bf375a561884 |
GradSurgeryGradSurgery is originally for balancing multi-task learning. We can easily apply GradSurgery for auxiliary learning by specifying a task as the target task and the others as the auxiliary tasks {{cite:dc7700d3d8724c934ae603d5877817b6def04915}} replaces {{formula:849e0b86-0b52-4dcd-a64b-7809380ab8fd}} by its... | m | c3a7c020574dd40a0e50b6585216f7ce |
Based on the fact that tremendous progress has been witnessed in the deployment of QKD, practical secure QKD protocols are becoming an important research direction in quantum communication since they are relevant for the practical deployment of QKD . In this study, we propose an FP-TFQKD protocol using coherent states ... | d | 259d369558b512ec4d63bf0d8463b7ec |
Contributions: In this work, we introduce IMEA, an Informed Multi-context Entity Alignment model, to tackle the above obstacles.
Motivated by the strong expressiveness and representation learning ability of Transformer {{cite:37e5ac543402c90c39a03a7b2ce0b466d1f1c2cd}},
we design a novel Transformer architecture to enco... | i | 3757a3f292843b87f0f4dec99aecdf78 |
As the remnant radio galaxy ages, its electrons will lose so much energy that their spectra will eventually steepen. This steepening would reach to a point where they do not emit significant radiation at high frequencies, instead, the radio emission may only be observable at low frequencies.
However, {{cite:08b609c995c... | d | 9300e1004273b1984f04db91fc926a67 |
Quantum state engineering has been a subject of increasing interest to
construct various novel nonclassical states in quantum optics and quantum
information processing{{cite:f7d418c576ae65923e6acc8016ee8244fe4cd0a6}}, {{cite:f9931816686afa7737a00a1d61d572b0f3b52eb5}}. From a theoretical point of view, the
simplest way ... | i | 072df33578f4559a002ab3862989da64 |
The recent discovery of superconductivity in the strongly correlated system UTe{{formula:62534752-53a1-4442-8d98-50d1729123b9}} has sparked enormous interest {{cite:c5801f2decb6ac278c4ebccf1d21b0fea819f302}}, {{cite:84790b1c90101404a4307d891f8fd528c3f6e61d}}. This orthorhombic compound is a paramagnet with anisotropic... | i | 09e5d1440f75a055631e17e325a3e665 |
The existing methods for robust recommendation {{cite:3e1b7d6ca3a4c7249eccfebcc3046f43d9c998d1}}, {{cite:a9d170a79896474940380d8471a199efc832cc63}}, {{cite:7540ddfde83fe59ef14cbb4edca986413331392f}}, {{cite:db35307c13c0a5dbd9ce9d1a18eb577fc91d4bf6}} often inject extra noise to training data or model parameters to deal ... | m | 6603286b553e52fe676caf4e9a4d77cc |
For UG2-2.1, we report baseline results from a total of six AR models including: I3D {{cite:0b404d195caed0aaf417cb945573a01f3745963c}}, 3D-ResNet-50 {{cite:a661bb81ee37fa2e16cb8343adeb7dbc2ba923be}}, 3D-ResNeXt-101 {{cite:a661bb81ee37fa2e16cb8343adeb7dbc2ba923be}}, TSM {{cite:456c24509ef09f3e5b1162b1a724dd14fb6a7f05}},... | r | e9b9070d4227fb2a179a87848c8152ff |
Reinforcement learning (RL) algorithms have achieved state-of-the-art performance
in many fields including locomotion control {{cite:1a14dc407d1521a112c4097cd81955cfc8a6b57e}}, {{cite:246e45e7f8d090b711f2223b8cb22cc0e7304847}}, autonomous driving {{cite:c8acd276ed8f3e0e7fec6ab9168c07b9564f422f}}, {{cite:a813d5f485e9d6c... | i | 17d7b97734a55b91bfc31e60062865a9 |
As mentioned above, operational semantics {{cite:3723545b5f217b77c6aeb9e39edc345734c07523}}, {{cite:623536612797549bc1d50cd515071570a9d8fce0}} captures the meaning of an executable program by the environment transitions according to the instructions on an abstract machine. To be more concrete, we illustrate our motivat... | m | 068f55457c9dbbd6ffc9f61e5909d4ec |
Due to lack of space, state-of-the-art in meta-classifiers has been omitted from the literature review. However, note that this work is focused on zero-shot learning. Hence, five standard meta-classifiers have been tried: (A) Meta-Decision Tree MDT {{cite:0d1447335accdacc0d12a980440df07860dfce6b}}, (B) deep neural netw... | m | 6bf56d4d9fe3b85e5793b64bed152095 |
The two temperature model represents the next logical step on expanding from a one temperature model, and is motivated by previous literature {{cite:9114142054a4a352bc1b512f9f962382cfe2821a}}, {{cite:0d2ce0d8cbb11c1f61aee3f2248193bf353a4834}}. However, the particular temperature values found using the two temperature m... | d | ee8f006e2be48ccb5e1bcf4fce932bb6 |
First, in Sec. , we consider the absolute retracts of bipartite graphs and some important subclasses of the latter. We observe that in the square of such graph {{formula:a016faf5-2800-46cc-b665-31afbc9ca6ee}} , its two partite sets induce Helly graphs. This result complements the known relations between Helly graphs a... | r | 9fcea7a764773055f0c661d4691bfa87 |
On the mathematical side, the problem of SMEFT tree-level matching
up to dimension-six terms was completely solved in Ref. {{cite:88a39565a1da2754c285137822d27b1d1f96d8df}},
which provided the complete dictionary of SMEFT contributions for
all possible tree-level mediators. The key observation behind this
work is that,... | m | 689470b9d2f73a00c1c2fa4852c1add1 |
Accelerated methods for multiobjective optimization are not sufficiently discussed from a theoretical point of view in the literature yet. In {{cite:fa75f6507917cd90744bcb6a44454d4a21b17442}} El Moudden and El Moutasim propose an accelerated method for multiobjective optimization which incorporates the multiobjective d... | m | 3655765598e91ee0de80d77944877429 |
Rather than plotting the full distribution, it is also common {{cite:3a26882bccd4b03c26ec340420e940453d5624b4}}, {{cite:aa8ebe20641a15f7a0b73d3046df8daa5e9b7e06}}, {{cite:c9e2ed80c91068a3a5c792ba0987ea552fb7009f}} to compare the posterior with the ground truth parameters of synthetic data.
Such a comparison does not re... | m | 3cc8572fc1d3c7f7918cc9ee3f262963 |
Although a promising quantitative explanation for the researched phenomenon is developed, it is yet indiscreet to conclude that EpiRank is by any means a sufficiently accurate index of urban epidemic hazards. The current dynamic network model draws little besides the two aspects, urban population and inter-city transpo... | d | 61fe2bb00d3f0bdc338f6d8926f6a8d6 |
where {{formula:0cb98e56-797f-4f65-8785-bf7865ed8aaf}} and {{formula:1de47bb2-d49f-492a-b432-72b1bf10a004}} are their 2nd Fourier coefficients, i.e., total transverse rapidity flow and radial flow {{cite:ba44e304cf0f7c5c383561aca6e1503d92725d56}}, {{cite:2197f44d75a453d379ccac63d4296dde6137acb4}}, {{cite:8b7c854f8498... | m | 7581a9f1ab42d5726837d835755b3669 |
Other methods such as {{cite:1c64cf815859175352e445af4e35b61fa15a4589}}, {{cite:4cffc18ef87f1c7d1fb4f573502fc4e5c3204bc9}}, {{cite:f97bf0cb8a3d6b20827693037d00d60f57310273}}, {{cite:dc2603b33ea3c91cbd4203d41731076a60940ad8}}, {{cite:ea8f2b59b1209d2ad4345c4ad686229c13625772}}
and {{cite:f6b9a990b03ab7dbd8cbc454e19d9fd3c... | m | 8575c205bb1961400f0046b39ba5ac6d |
Each method described above was applied to deep CNNs tested on visual classification. Models were trained on the MNIST handwritten digits dataset, the CIFAR-10 image set, and the ImageNet dataset. On all three datasets, we implemented relatively simple baseline architectures following the basic paradigm established by ... | r | 0ed5a41f9452166a3593654740e5a8d3 |
Mahalanobis Detector
The above three methods are based on the confidence. Another approach is to formulate the problem as unsupervised anomaly detection.
{{cite:dd171df2b7cea410f8c15cfdcd420e4753ed06d9}} proposed to model the distribution of intermediate layer's activation by a Gaussian distribution for each class bu... | m | 21a53e0527c549cebb7fa992cc1bce61 |
We train our method on ShapeNet {{cite:c2d85eb027d3f6afe5ef451801689cd27e9716bd}} data set. In order to achieve better generalization we use objects from all classes. In each step the object class and corresponding object from it selected randomly. We follow this strategy for both pre-training and training phases. To a... | r | e65ea8b2c3c0bbdca9d437a5628a102d |
For evaluation we use BLEU {{cite:ccc6c9928bb67ccb40c699ff7bf95ce97fd9eed2}}, ROUGE {{cite:0247153b7479e978e481dbe573bb7c6fdafa55d1}} and METEOR {{cite:9cb180b98b16dfea3b7fa2fa905d3674302058c3}}. For BLEU, we report BLEU-1,2,3,4 scores and as ROUGE score we report the ROUGE-L F1 score. These metrics allow us to directl... | d | fbafceb4ba1aa1d85316ef99e1dcde3f |
When we compare the RS and NC network methods to the random guess method, we find that all network methods, except RESCAL and DEDICOM, are significantly (p-value{{formula:e97dff84-c470-409f-8224-42e296a96688}} 0.05) more accurate in terms of all evaluation measures (Figures REF , REF and Supplementary Figure S5). The ... | r | 148d631ce4677e602c2e916823118e63 |
The aim of this study was to evaluate the bias of face-based BMI prediction models across four gender-racial groups (Black Females, Black Males, White Females, and White Males).
Experimental results suggested performance differential of facial analysis-based BMI prediction tools.
However, in-contrary to bias analysis o... | d | 897e67d711ec7af6e4dbb306dd414050 |
Films of known thickness were measured by compressing powder
within apertures between the KBr discs. Apertures were made in the
{{formula:0d08f786-9cd2-4147-9af2-5dedc5cd0e75}} -thick microphone foils measured by {{cite:1d4d4a1259458c3adb34b5e4e4d93b8b4a58fdb5}}
and with with 6 -thick tin gaskets used for IR cells (th... | m | d0f6841e4b575ad43f0fbed00a0c2c2b |
We compare our proposed algorithms CORAL-CEM and MMD-CEM to ERM {{cite:d0073302f022bb70ce844615a60473f6de7479de}}, IRM {{cite:063aea776f3093a029aed156c4563e4db2657423}}, IB-ERM {{cite:c56a5eabb1c2f8f19cb706a8726fb1c438fb8282}}, IB-IRM {{cite:c56a5eabb1c2f8f19cb706a8726fb1c438fb8282}}, MMD-IRM {{cite:d84d2a13fb4e9764f50... | m | 13f5386f6b4d3d53d588bf57687a9042 |
Real-world populations.
In Table REF , we report values of the rescaled average RSE distance {{formula:9b162c30-5d96-471c-b6b5-46699f760d82}} for various real-world networks together with their degree-heterogeneity {{cite:174ee657f82ca0271f6d18aaa35b92d6758a5ea6}} and clustering coefficients {{cite:56cac892c2d9d5a7417... | r | 6abea548571641c6979eed39d91297b1 |
We randomly sample {{formula:bc0678e8-b182-4277-8a84-cfe84610ddc1}} sentences from the Toronto BookCorpus {{cite:b08c1b33fd93f816df2fd79b0adb85f6b4eaa4dd}}—a collection of English books of various genres—for computing Mantel tests using Levenshtein distance, both raw and normalized, as the textual distance.
We repeat ... | m | c6ddcba6b895e123b86a43b1e46071ab |
The advantage of GPS is twofold. First, the sampling is guided by the designated adaptive sampling policy, enabling a more effective message propagation than GraphSAGE, whose node representations are generated through random sampling of neighborhoods of each node. Other importance-based sampling works adopt a different... | d | 6339e01be6bc980068e99e605c65c033 |
Unlike the previous research that combines two independently trained systems in parallel, in this paper, we propose a two-pass hybrid and E2E cascading (HEC) framework as shown in Fig. REF . A conventional hybrid system is built as the first pass model, which outputs segmented audio and its corresponding N-best hypothe... | i | 9327f34d150e76a2bca4e1b4febd230d |
While our framework leverages the work of {{cite:6a2df1e67f39c57c43133c7c8c9790c0f47cc04b}} to enter the SNN domain, this work also introduces novel results and further innovations. In the {{cite:6a2df1e67f39c57c43133c7c8c9790c0f47cc04b}} experiments, results were shown for networks with plastic and neuromodulated syna... | d | bf92606d3a14c93eccf423ce5c42d8bf |
Discriminative patches are the patches that were correctly classified by the model with high confidence. The softmax probabilities of the predictions from the classification model were used to obtain the confidence scores. Despite achieving better performance, modern neural networks tend to be overconfident. This can b... | m | cbd602408b64f2f8da5bda9ee32974c5 |
Our interest in the graph {{formula:da133439-b270-4145-a3fc-366918be5846}} was triggered by the following question, which is related to Schramm's locality conjecture, see {{cite:e8fac29c329d592198c72f4cae99e15a418454c4}} and also {{cite:d33062ab07e5c34ca384cf430f1a7025f78282e4}}, {{cite:a27ae3e677b899de7bbc0d16eba61f6... | r | 69eababde1e9f3187d4a7fc4c86dcf24 |
Recently, the success of deep learning in the fields of image and activity recognition has motivated its application to gait recognition. The basic feature representations of these methods are also gait templates. {{cite:d4a2646992cff49dc06c29b7b98585f582a7e2bb}} employed a CNN (Convolutional Neural Network) to define ... | m | 6ae272e9bee7a1539c76ea794f5f1967 |
The second set of experiments compared our proposed framework with other baselines in the setups of linear evaluation, fine-tuning and few-shot transfer learning. In addition to our model (SSL with LPA), we studied 4 baselines: SSL with LAA, supervised models trained without augmentations, supervised models trained wit... | m | 6695ddf22d22a8464b7a978539f575e0 |
However, the applicability of our result does not stop at (syntactic) extensions of our framework, as it applies to arbitrary query languages and querying formalisms of different types. In particular we would like to stress the relationship to the very comprehensive existential rules fragment of bounded treewidth sets ... | d | f14b154fb10fde7e74f6be991c319c09 |
In the present work, only the variable `sex' was considered as a sensitive attribute. However, in many practical applications, multiple sensitive variables may need to be considered. For example, the models may be considered fair with respect to gender but still make discriminatory decisions with respect to race. Likew... | d | 1b202b44532b0a55f15a529acaefb406 |
In the literature of multitask learning, {{cite:e02e18cc150d927ce02dba71ad6324c4c7c58821}} showed that the excess risk, namely the difference between the learned parameters and the optimal ones, scales as {{formula:3c448d2a-70c4-44b7-b1ad-cecc67878f17}} , where {{formula:65e34783-12c1-4a9d-a14e-841458e78f6b}} is the n... | d | 226f673a8f93c073da2b1abd5e409efe |
More formally, in the centralized setting we present a {{formula:bc17c539-a638-4934-a6e1-7ccabb79295d}} -approximation algorithm with summary size {{formula:515aa54e-92eb-448e-a649-41e08f507bb6}} . In the streaming setting we provide a {{formula:318fd7b3-322c-476d-b92e-68dd84cc1ca4}} -approximation algorithm with summa... | r | 8a23c6f34ed4be134afebf92816ee869 |
We use Ramulator {{cite:ee6a660b92b0191da1e823b17674ddf880490b88}}, {{cite:151b7fedf88bb1ea365496283dcfaa6f085fbedb}}, a cycle-accurate
DRAM simulator with a simple core model and a system configuration as listed in
Table REF , to implement and evaluate the RowHammer
mitigation mechanisms. To demonstrate how the per... | m | 38eae5b666e30b8158819313110aa8e4 |
The value of DIC can be computed directly from the Markov chains generated with MontePython.
For values {{formula:357adfd4-3b0d-409c-9466-3e964358fba4}} we would conclude strong evidence of the RRVM's as compared to the {{formula:f451039c-3643-45fd-b3a6-4df5c2b78e3e}} CDM, and for {{formula:db4ffc77-032f-490f-9ebf-9d6... | r | 6ecf8710790404223def93db74951e51 |
Let {{formula:fde29690-a704-4f57-8e9c-2b4a182386f7}} denote the set of digraphs.
Various important results in graph theory have been obtained by considering some functions
{{formula:ba25ae5a-a6bb-41bd-883d-db22ebed987c}}
or {{formula:4f0ee8e8-dbd8-4d06-a3f2-21549a5e19ba}} called
operations
(here each {{formula:3251d... | i | 4a91dcac623211d02a3e0122519437c5 |
We train our model with both detection and segmentation heads to verify the learning ability of our model for multiple tasks, and the results are shown in Tab. REF .
While comparing different BEV encoders under same settings, BEVFormer achieves higher performances of all tasks except for road segmentation results is co... | r | 0edb8b80d97f5b744c9ed4d45164fc61 |
In this section, we present numerical results to verify the performance of the proposed transmission scheme based on Bayesian optimization. The BS is equipped with {{formula:b66ac639-e77a-4d74-a341-099f61cf2f5c}} antennas. The number of elements in RIS is {{formula:4b756003-cb4f-4845-a997-57d49a568a33}} if there is n... | r | 53cd9e03c419cba5682a7eeabeaa4f68 |
Recent GANs {{cite:18a3ce4cab855ff1df66d798518883399d952973}}, {{cite:5217899c5ed560ff00e61a9b20fec7df04a1e65d}} demonstrated that synthetic images can be generated with very high quality. This motivates research into embedding algorithms that embed a given photograph into a GAN latent space. Such embedding algorithms ... | i | c797876482fbe1ee14bcd1249e8afb53 |
In Fig. 2, the critical temperature of the three dimensional
ternary alloy model has been plotted as a function of {{formula:4b4f5381-6389-48d6-b557-a91171916d02}} for
various values of {{formula:c705d3e7-6a66-446a-8e90-7cec0557a2ba}} when {{formula:58edf2ca-e769-4b8a-96a9-5e8aa4a5b10b}} . It can be seen from Fig. 2
... | r | d4c8e6672ac8261ae2c7caa6f4fe862f |
The choice of an entropic penalty does not require specifying a
reference measure {{formula:682c17c7-62db-4549-86ed-4802fa5a47e5}} and connects (REF ) with the
classical maximum entropy methods {{cite:481e174a7fa88bba1b9a625c804967a261b6e1ba}}. Indeed,
(REF ) is the Lagrangian associated with the following primal
prob... | m | a0f6426ee0e85e3653c139be86a34977 |
We also note that this paper is limited by the model of opinion dynamics. Experimental research has shown that exposure to differing opinions may increase polarization {{cite:079b168d731c892dc58c853959b45f85732a5cfd}}. There are recent extensions to the Friedkin-Johnsen model that incorporate opinion repulsion, such as... | d | fd658420cd596ad4fe229509237e4ea7 |
We first used two baselines, by following {{cite:07722ec90264cc9d36317a283d87ee34627a84f1}}, i.e. the mean of training drug response values as a prediction for the unobserved drug responses, where we considered 1) cell-line specific mean (cls-mean) and 2) overall mean (all-mean).
We then used two state-of-the-art machi... | m | e6cc26f347380c0f29512e26cd360b38 |
Phase separation organising chemical reactions was suggested as an important concept to understand cellular biochemistry {{cite:1aaef3e956fb2a758ae39af05ea72b5a9e1c6e3d}}, {{cite:f88335f33bcbd45035d0d6a71319eed9cca4d2ba}}, {{cite:6085950d6310318de16399fe3ab73418a2e5af1f}}, {{cite:89922af63daf7ea3593ea50c2f99e84d95a77ef... | d | 7c806e7eaa649f0a4b7293d477ff116e |
The second (“hybrid") corpus was generated by combining the authentic content and the machine generated content. For this task we utilized some of the latest abstracts from Artificial Intelligence domain in ArXiv. Each hybrid abstract is made of 4 parts. The initial content is extracted from an original abstract up to ... | m | 1f552655bcbceb917140b8b29c1ad1e8 |
Fig REF displays the density wake created by our fiducial single perturber after it is turned on at {{formula:be43c3d1-1d2a-46ad-af5f-ccaba03d9e0e}} . Results are shown both in the Lippmann-Schwinger approach ({{formula:a97e5eb2-635a-4494-a28a-587dc74a827b}} , left panel) and in the Madelung approach ({{formula:b5d526... | r | 1dd12c69e8cc7e9de0cf4ff05771c4b7 |
Although slow-roll is almost an inevitable requirement of inflation, the observed small temperature fluctuation in the universe (or, more precisely the small CMB amplitude) is not. Slow-roll inflation could have explained homogeneity, causality, and the quantum origin of density perturbations without predicting the sma... | i | 92ac6239fec2a7c577f93253d62e042b |
Similarly, the SFR is very low ({{formula:c9b78335-664a-4c0b-b4b4-4cdbe0af83ec}} 0.1 M{{formula:96130fb8-e15f-4879-a9be-0a9c8773cd03}} yr{{formula:6bf3976b-6902-496b-87a3-1a0b82f4efde}} ) for half of the sample with an average value of {{formula:bf48274d-3a09-4ff5-b862-c7de836bbed0}} M{{formula:42a98b21-c98d-4fb7-8f... | d | 8f4d3de4105667da987a0c55b8d019bf |
Why should one expect eq:srequals2ew to be true? An intuitive proof, also discussed in Ref. {{cite:f8bbb068268b51c78a985a6543b8253711e9b538}}, is that {{formula:fadacbde-6413-4f49-af15-cd35aafc1eff}} is dual to two copies of the original entanglement wedge, glued along the RT surface as depicted in fig:EW. The entropy... | i | 0c117594615802e2fa3824333cc614d8 |
Hypergraph, a generalization of graph containing edges that can be incident to more than two vertices, serve as a natural tool to model such complex and high order relationships. For example, In a co-citation hypergraph, hypernodes represent papers and hyperedges mean the citation relationships. The ubiquity of complex... | i | 6352526c499e22d5eddd398800908d91 |
For the most popular distance metrics—the Euclidean distance ({{formula:b0075072-7161-4e27-805f-88920d54534c}} -norm) and Manhattan distance ({{formula:59ffd5a3-bf94-4b7e-80f9-6c141d4db061}} -norm)—classic dimension-reduction (sketching) provide very efficient distance oracles {{cite:416f1d22b7984440c958862886a3e2fa343... | i | c863f917cab2a0f7108c64a3b480f724 |
Furthermore, four topics beyond the scope of this paper are worth further
investigation. In this paper, our proposed
algorithm is based on two step estimation under piecewise constant
proportional hazard model assumption. Proposing an efficient
sampling algorithm without Laplace approximation is an important future wor... | d | 80ea3778acbe6343186f33ce09495a36 |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.