text_with_holes stringlengths 111 3.44k | text_candidates stringlengths 46 1.43k | A stringclasses 6
values | B stringclasses 6
values | C stringclasses 6
values | D stringclasses 6
values | label stringclasses 4
values |
|---|---|---|---|---|---|---|
Of course, the numerical scheme and the estimates developed in Section 3.1 hold. However, several simplifications are possible when the coefficients have low-contrast, leading to sharper estimates. <|MaskedSetence|> <|MaskedSetence|> Also, our scheme is defined by a sequence of elliptic problems, avoiding the annoyan... | **A**: We remark that in this case, our method is similar to that of [MR3591945], with some differences.
**B**: We had to reconsider the proofs, in our view simplifying some of them.
.
**C**: First we consider that T~~𝑇\tilde{T}over~ start_ARG italic_T end_ARG can be nonzero.
| ACB | ABC | ACB | ACB | Selection 3 |
<|MaskedSetence|> The best feature is related to sentiment polarity scores. There is a big difference between the sentiment associated to rumors and the sentiment associated to real events in relevant tweets. <|MaskedSetence|> Furthermore, we would expect that verified users are less involved in the rumor spreading. ... | **A**: In specific, the average polarity score of news event is -0.066 and the average of rumors is -0.1393, showing that rumor-related messages tend to contain more negative sentiments.
**B**: For analyzing the employed features, we rank them by importances using RF (see 3).
**C**: Also interestingly, “IsRetweeted” ... | BAC | BAC | BCA | BAC | Selection 4 |
For the evaluation, we shuffle the 180 selected events and split them into 10 subsets which are used for 10-fold cross-validation (we make sure to include near-balanced folds in our shuffle). <|MaskedSetence|> Furthermore, we implement the neural network with TensorFlow 121212https://www.tensorflow.org/ and Keras13131... | **A**: For the experiments, we implement the 3 non-neural network models with Scikit-learn library111111scikit-learn.org/.
**B**: The output of the embedding layer are low-dimentional vectors representing the words.
**C**: To avoid overfitting, we use the 10-fold cross validation and dropout for regularization with d... | ABC | ABC | ABC | CAB | Selection 2 |
Evaluating methodology.
For RQ1, given an event entity e, at time t, we need to classify them into either Breaking or Anticipated class. <|MaskedSetence|> <|MaskedSetence|> For GoogleTrends, there are 2,700 and 4,200 instances respectively. We then bin the entities in the two datasets chronologically into 10 differen... | **A**: We select a studied time for each event period randomly in the range of 5 days before and after the event time.
**B**: We set up 4 trials with each of the last 4 bins (using the history bins for training in a rolling basic) for testing; and report the results as average of the trials..
**C**: In total, our tra... | ACB | ACB | ACB | BAC | Selection 3 |
Table 1 shows basic patient information. Half of the patients are female and ages range from 17 to 66, with a mean age of 41.8 years. <|MaskedSetence|> <|MaskedSetence|> <|MaskedSetence|> In terms of time since being diagnosed with diabetes, patients vary from inexperienced (2 years) to very experienced (35 years),... | **A**: The mean BMI value is 26.9.
**B**: Only one of the patients suffers from diabetes type 2 and all are in ICT therapy.
**C**: Body weight, according to BMI, is normal for half of the patients, four are overweight and one is obese.
| CAB | CAB | BAC | CAB | Selection 1 |
To restore the original image resolution, extracted features were processed by a series of convolutional and upsampling layers. <|MaskedSetence|> (2018); Liu and Han (2018), but we argue that a carefully chosen decoder architecture, similar to the model by Pan et al. (2017), results in better approximations. Here we e... | **A**: The outputs of all but the last linear layer were modified via rectified linear units.
**B**: This setup has previously been shown to prevent checkerboard artifacts in the upsampled image space in contrast to deconvolution Odena et al.
**C**: Previous work on saliency prediction has commonly utilized bilinear ... | CBA | CBA | ACB | CBA | Selection 1 |
Our strongest positive result about the approximation of the locality number will be derived from the reduction mentioned above (see Section 5.2). <|MaskedSetence|> <|MaskedSetence|> Firstly, ruling out simple strategies is a natural initial step in the search for approximation algorithms for a new problem. Secondly... | **A**: This is mainly motivated by two aspects.
**B**: It may seem naive to expect new approximation results for cutwidth in this way, but, as mentioned in the introduction and as shall be discussed in detail in Section 6, approximating the cutwidth via approximation of the locality number may be beneficial for cutwid... | CAB | CAB | CAB | CAB | Selection 4 |
Oh et al. (2015) and Chiappa et al. <|MaskedSetence|> <|MaskedSetence|> <|MaskedSetence|> (2017), however we focus on using video prediction in the context of learning how to play the game well and positively verify that learned simulators can be used to train a policy useful in original environments.. | **A**: (2015) and Chiappa et al.
**B**: Impressively, in some cases the predictions maintain low L2subscript𝐿2L_{2}italic_L start_POSTSUBSCRIPT 2 end_POSTSUBSCRIPT error over timespans of hundreds of steps.
As learned simulators of Atari environments are core ingredients of our approach, in many aspects our work is m... | CBA | CBA | ABC | CBA | Selection 1 |
<|MaskedSetence|> WorkPartner [8] demonstrated its capability to seamlessly transition between two locomotion modes: rolling and rolking. The rolking mode, a combination of rolling and walking, empowered WorkPartner to navigate with enhanced agility. <|MaskedSetence|> However, it’s noteworthy that Gorilla only has wa... | **A**: This feat was accomplished through the implementation of devised criteria that took into account a comprehensive analysis of energy utilization, wheel slip percentage, and the intricate dynamics between the wheels and the demanding terrain.
**B**: The threshold values for locomotion transition criteria were est... | BAC | CAB | CAB | CAB | Selection 3 |
It should be fairly clear that such assumptions are very unrealistic or undesirable. Advice bits, as all information, are prone to transmission errors. <|MaskedSetence|> <|MaskedSetence|> For a very simple example, consider the well-known ski rental problem: this is a simple, yet fundamental resource allocation, in w... | **A**: In contrast, an online algorithm that does not use advice at all has competitive ratio at most 2, i.e., its output can be at most twice as costly as the optimal one..
**B**: In addition, the known advice models often allow
information that one may arguably consider unrealistic, e.g., an encoding of some part of... | BCA | ABC | BCA | BCA | Selection 1 |
<|MaskedSetence|> For classifiers supporting incremental classification, like SS3 or MNB, only a small vector needs to be stored for each user. For instance, when using SS3 we only need to store the confidence vector303030In case of ADD, a 2-dimensional vector. of every user and then simply update it as more content i... | **A**:
It is worth noting that the difference in terms of space complexity is also very significant.
**B**: However, when working with classifiers not supporting incremental classification, for every user we need to store either all her/his writings to build the document-term matrix or the already computed document-t... | ABC | ABC | ABC | BCA | Selection 1 |
Fig. 12 presents the sketch diagram of a UAV’s utility with power altering. <|MaskedSetence|> When other UAVs’ power profiles are altering, the interference increases and the curve moves down. The high interference will reduce the utility of the UAV. Fig. 12 also shows that utility decreases and increases with power i... | **A**: Small and large power both provide high utilities, which is because small power will save energy and large power will increase SNR.
**B**: The altitudes of UAVs are fixed.
**C**: For the sake of enlarging the global utility, the largest power is not the optimal strategies for the whole UAV ad-hoc network.
| BAC | BAC | ACB | BAC | Selection 2 |
The sources of DQN variance are Approximation Gradient Error(AGE)[23] and Target Approximation Error(TAE)[24]. <|MaskedSetence|> This type of variance leads to converging to sub-optimal policies and brutally hurts DQN performance. The second source of variance Target Approximation Error which is the error coming from... | **A**: Dropout methods have the ability to assemble these two solutions which minimize different source of variance.
**B**: In Approximation Gradient Error, the error in gradient direction estimation of the cost function leads to inaccurate and extremely different predictions on the learning trajectory through differe... | BAC | BAC | BAC | ABC | Selection 3 |
Figure 5: Top: An illustration of the SegNet architecture. <|MaskedSetence|> Bottom: An illustration of SegNet and FCN (Long et al., 2015) decoders. a,b,c,d𝑎𝑏𝑐𝑑a,b,c,ditalic_a , italic_b , italic_c , italic_d correspond to values in a feature map. SegNet uses the max-pooling indices to upsample (without learning)... | **A**: There are no fully connected layers, and hence it is only convolutional.
**B**: Note that there are no trainable decoder filters in FCN (Badrinarayanan et al.
**C**: FCN upsamples by learning to deconvolve the input feature map and adds the corresponding encoder feature map to produce the decoder output.
| ACB | CBA | ACB | ACB | Selection 3 |
<|MaskedSetence|> (2014), each dataset is split into a training and a test set using a 50/50 split while maintaining the label distribution. <|MaskedSetence|> We evaluate the training with 5555, 10101010, 20202020, and 50505050 examples per class.
In contrast to Fernández-Delgado et al. (2014), we extract validation ... | **A**: Afterward, the number of training examples is limited to nlimitsubscript𝑛limitn_{\text{limit}}italic_n start_POSTSUBSCRIPT limit end_POSTSUBSCRIPT examples per class.
**B**: For some datasets which provide a separate test set, the test accuracy is evaluated on the respective set..
**C**: Following Fernández-D... | CAB | CBA | CAB | CAB | Selection 1 |
Broadly speaking, our work is related to a vast body of work on value-based reinforcement learning in tabular (Jaksch et al., 2010; Osband et al., 2014; Osband and Van Roy, 2016; Azar et al., 2017; Dann et al., 2017; Strehl et al., 2006; Jin et al., 2018) and linear settings (Yang and Wang, 2019b, a; Jin et al., 2019;... | **A**: (2019); Ayoub et al.
**B**: (2020); Zhou et al.
**C**: (2019).
| BCA | ABC | BCA | BCA | Selection 1 |
We thank Prof. Henry Adams and Dr. Johnathan Bush for very useful feedback about a previous version of this article. <|MaskedSetence|> <|MaskedSetence|> Michael Lesnick for explaining to us some aspects of their work. We thank Dr. Qingsong Wang for bringing to our attention the paper [76] which was critical for the p... | **A**: Finally, we thank Dr.
**B**: We also thank Prof.
**C**: Mikhail Katz and Prof.
| BCA | BCA | BCA | ACB | Selection 2 |
After choosing a projection, users will proceed with the visual analysis using all the functionalities described in the next sections. <|MaskedSetence|> The top 6 representatives (according to a user-selected quality measure) are still shown at the top of the main view (Figure 1(e)), and the projection can be switched... | **A**: In the case of the selection-based quality, we aggregate only over the selected points to reach the final value of the quality measure, which is then used to re-rank the representatives.
.
**B**: However, the hyper-parameter exploration does not necessarily stop here.
**C**: After selecting these points, the l... | ABC | BCA | BCA | BCA | Selection 4 |
Good comparisons are crucial for new proposals: The lack of fair comparisons is another important drawback of many proposals published to date. <|MaskedSetence|> These algorithms have been widely surpassed by more advanced versions over the years which, so obtaining better performance than naive version of classical a... | **A**: When new algorithms are proposed, unfortunately, many of them are only compared to very basic and classical algorithms (such as GA or PSO).
**B**: We encourage researchers to increase the algorithms used in their experimental section, including more competitive or state-of-the-art algorithms: until they are pro... | ACB | CBA | ACB | ACB | Selection 4 |
<|MaskedSetence|> Instead, DEC and SpectralNet work better on the large scale datasets. Although GAE-based models (GAE, MGAE, and GALA) achieve impressive results on graph type datasets, they fail on the general datasets, which is probably caused by the fact that the graph is constructed by an algorithm rather than pr... | **A**: Classical clustering models work poorly on large scale datasets.
**B**: In particular, AdaGAE is stable on all datasets.
.
**C**: The adaptive learning will induce the model to exploit the high-level information.
| CBA | ACB | ACB | ACB | Selection 3 |
Ingress filtering. To enforce ingress filtering the networks should check the source address of an inbound packet against a set of permitted addresses before letting it into the network. Otherwise, the attackers using spoofed IP addresses belonging to the network can trigger and exploit internal services and launch at... | **A**: Enforcing ingress filtering is therefore critical for protecting the networks and the internal hosts against attacks.
**B**: For instance, by spoofing internal source IP addresses the attackers can obtain access to services, such as RPC, or spoofed management access to networking equipment [RFC3704], the attack... | BAC | ABC | BAC | BAC | Selection 1 |
This paper builds upon previous work with this dataset [7], which used support vector machine (SVM) ensembles. <|MaskedSetence|> Context-based learning is then introduced to utilize sequential structure across batches of data. <|MaskedSetence|> <|MaskedSetence|> Thus, emulation of adaptation in natural systems leads... | **A**: The context model has two parts: (1) a recurrent context layer, which encodes classification-relevant properties of previously seen data, and (2) a feedforward layer, which integrates the context with the current odor stimulus to generate an odor-class prediction.
**B**: The results indicate improvement from tw... | CAB | CAB | CAB | BCA | Selection 1 |
<|MaskedSetence|> This culminated in constructions to present free groups of arbitrary rank as automaton groups where the number of states coincides with the rank [18, 17]. While these constructions and the involved proofs are generally deemed quite complicated, the situation for semigroups turns out to be much simple... | **A**:
There is a quite interesting evolution of constructions to present free groups in a self-similar way or even as automaton groups (see [15] for an overview).
**B**: In fact, the construction to generate these semigroups is quite simple [4, Proposition 4.1] (compare also to 3).
**C**: The same construction can ... | ABC | ABC | BAC | ABC | Selection 4 |
While our results indicate that current visual grounding based bias mitigation approaches do not suffice, we believe this is still a good research direction. However, future methods must seek to verify that performance gains are not stemming from spurious sources by using an experimental setup similar to that presented... | **A**: (2019a).
.
**B**: Finally, we advocate for creating a dataset with ground truth grounding available for 100% of the instances using synthetically generated datasets Kafle et al.
**C**: (2018); Acharya et al.
| BCA | BAC | BCA | BCA | Selection 1 |
<|MaskedSetence|> Simhashing is a hashing technique in which similar inputs produce similar hashes. <|MaskedSetence|> We then obtained a list of all pairs of similar documents based on a distance threshold (measured based on the number of differing bits) that was determined after manual examination of a number of pai... | **A**: We then filtered the duplicates based on a greedy approach retaining policies that were longer in length.
**B**: After creating shingles (Broder et al., 1997) of size three, we created 64 bit document Simhashes and measured document similarity by calculating the Hamming distance (Manku et al., 2007) between doc... | CBA | CBA | CBA | CBA | Selection 3 |
The process of ensemble learning generates a solution space of models (Figure 1, curved green arrow) [47] and more investigations can be done to choose between the best and most diverse models of an ensemble (Figure 1, curved red arrow). <|MaskedSetence|> automated approaches are essential concepts when developing a V... | **A**: The careful design, choice, and arrangement of these aspects and the balance between human-centered vs.
**B**: If a VA system supports asynchronous and/or synchronous communication, an individual expert can share his/her knowledge with the others, which could lead to a more desirable outcome.
.
**C**: Building... | CBA | ACB | ACB | ACB | Selection 4 |
Data Quantity. <|MaskedSetence|> <|MaskedSetence|> In Persona, the C Score and BLEU of MAML outperform baselines on 50-shot and 100-shot settings, but on 120-shot setting, the BLEU of MAML is lower than Transformer-F. <|MaskedSetence|> This finding is in line with the mechanism of MAML. MAML finds a sensitive parame... | **A**: In Weibo, FewRel and Amazon, the percentages that MAML outperforms the baselines by also decrease as the data quantity increasing.
**B**: In Persona, we evaluate Transformer/CNN, Transformer/CNN-F and MAML on 3 data quantity settings: 50/100/120-shot (each task has 50, 100, 120 utterances on average).
**C**: I... | ACB | BCA | BCA | BCA | Selection 3 |
When considering UAV communications with UPA or ULA, a UAV is typically modeled as a point in space without considering its size and shape. Actually, the size and shape can be utilized to support more powerful and effective antenna array. Inspired by this basic consideration, the conformal array (CA) [16] is introduce... | **A**: Regarding the mmWave CA, there are only a few recent works on the radiation patterns and beam scanning characteristics [20] and the performance evaluation of CA-based beamforming for static mmWave cellular networks [21].
**B**: Compared with surface-mounted multiple UPAs, a CA, conforming to the surface of a UA... | BCA | BCA | BCA | CAB | Selection 2 |
<|MaskedSetence|> <|MaskedSetence|> Specifically, the previous mean-field analysis casts SGD as the Wasserstein gradient flow of an energy functional, which corresponds to the objective function in supervised learning. In contrast, TD follows the stochastic semigradient of the MSPBE (Sutton and Barto, 2018), which is... | **A**: See also the previous analysis in the NTK regime (Daniely, 2017; Chizat and Bach, 2018a; Jacot et al., 2018; Li and Liang, 2018; Allen-Zhu et al., 2018a, b; Du et al., 2018a, b; Zou et al., 2018; Arora et al., 2019a, b; Lee et al., 2019; Cao and Gu, 2019; Chen et al., 2019a; Zou and Gu, 2019; Ji and Telgarsky, 2... | BAC | CAB | BAC | BAC | Selection 1 |
<|MaskedSetence|> (2019); Zhang et al. <|MaskedSetence|> We tested the performance of 6-layer models following the experiment settings of Zhang et al. (2020) for fair comparison. We adopted BLEU Papineni et al. (2002) for translation evaluation with the SacreBLEU toolkit Post (2018). <|MaskedSetence|> | **A**: (2020).
**B**:
To test the effectiveness of depth-wise LSTMs in the multilingual setting, we conducted experiments on the challenging massively many-to-many translation task on the OPUS-100 corpus Tiedemann (2012); Aharoni et al.
**C**: 111BLEU+case.mixed+numrefs.1+smooth.exp+tok.13a.
| BAC | BAC | BAC | CBA | Selection 1 |
<|MaskedSetence|> <|MaskedSetence|> <|MaskedSetence|> As benefits of the effective ordinal supervision and guidance of distortion information during the learning process, our approach outperforms Liao [12] by a significant margin, with approximately 23% improvement on PSNR and 17% improvement on SSIM. Besides the hi... | **A**: Specifically, compared with the traditional methods [23, 24] based on the hand-crafted features, our approach overcomes the scene limitation and simple camera model assumption, showing more promising generality and flexibility.
**B**:
As listed in Table II, our approach significantly outperforms the compared a... | ACB | BAC | BAC | BAC | Selection 4 |
<|MaskedSetence|> <|MaskedSetence|> The black-box model is motivated by data-driven applications where specific knowledge of the distribution is unknown but we have the ability to sample or simulate from the distribution. <|MaskedSetence|> Most prior work in this setting has focused on Facility Location [23, 24, 21,... | **A**: To our knowledge, radius minimization has not been previously considered in the two-stage stochastic paradigm.
**B**: Clustering is a fundamental task in unsupervised and self-supervised learning.
**C**: The stochastic setting models situations in which decisions must be made in the presence of uncertainty and... | BCA | ACB | BCA | BCA | Selection 3 |
<|MaskedSetence|> The co-existence of random graphs, subgradient measurement noises, additive and multiplicative communication noises are considered. Compared with the case with only a single random factor, the coupling terms of different random factors inevitably affect the mean square difference between optimizers’ ... | **A**: We firstly employ the property of conditional independence to deal with the coupling term of different random factors.
**B**:
III.
**C**: It becomes more complex to estimate the mean square upper bound of the local optimizers’ states (Lemma 3.1).
| BCA | ABC | BCA | BCA | Selection 3 |
<|MaskedSetence|> <|MaskedSetence|> Consequently, the accuracy of query answering of MuCo is much better and more stable than that of Mondrian and Anatomy. Besides, since the results of queries for MuCo are specific records rather than groups, the relative error rate of MuCo does not increase steadily with the growth... | **A**:
We observe that the results of MuCo are much better than that of Mondrian and Anatomy.
**B**: Therefore, differing from Mondrian and Anatomy, increasing the level of protection of MuCo has little influence on the query results.
**C**: The primary reason is that MuCo retains the most distributions of the origi... | ACB | CBA | ACB | ACB | Selection 4 |
Bells and Whistles. MaskRCNN-ResNet50 is used as baseline and it achieves 53.2 mAP. For PointRend, we follow the same setting as Kirillov et al. (2020) except for extracting both coarse and fine-grained features from the P2-P5 levels of FPN, rather than only P2 described in the paper. Surprisingly, PointRend yields 62.... | **A**: DCN and More Points Train.
**B**: Due to its superior performance, we only choose PointRend as ensemble candidates for the final submission..
**C**: The baseline trained on the official training set finally reaches 79.17 and 77.38 mAP on validation and testing set respectively, as shown in Table 1 and Table 3.... | ACB | ACB | BCA | ACB | Selection 4 |
<|MaskedSetence|> <|MaskedSetence|> One can apply the Cauchy-Schwarz inequality to show that our total variation bound implies that misspecification in Eq. (4) of Jin et al. is also bounded (but not vice versa). However, the regret analysis in the misspecified linear MDP of Jin et al. (2020) is restricted to static r... | **A**: (2020).
**B**: (1)).
.
**C**: The definition of total variation B𝐵Bitalic_B is related to the misspecification error defined by Jin et al.
| CAB | CAB | CAB | CAB | Selection 3 |
<|MaskedSetence|> <|MaskedSetence|> Most respondents encountered fake news on instant messaging apps compared to social media, and have reported the least trust in them. They have also rated the sharing of fake news to be a greater problem than its creation. These suggest that, in Singapore, communication with person... | **A**: Many studies worldwide have observed the proliferation of fake news on social media and instant messaging apps, with social media being the more commonly studied medium.
**B**: As an Asian country, Singapore tends towards a collectivist culture where emphasis is placed on establishing and maintaining relationsh... | ACB | ACB | ACB | ACB | Selection 1 |
<|MaskedSetence|> As discussed in previous sections, decentRL does not use the embedding of the central entity as input when generating its output embedding. <|MaskedSetence|> <|MaskedSetence|> Nevertheless, it still contains useful information for entity alignment. Additionally, decentRL benefits from concatenating... | **A**:
The performance of decentRL at the input layer notably lags behind that of other layers and AliNet.
**B**: The acquired information may not necessarily reside in the same dimension for a pair of aligned entities at this layer, which accounts for the comparatively lower performance of this layer.
**C**: Howeve... | ACB | ACB | CBA | ACB | Selection 1 |
Network architecture. The proposal network contains 2222 fully-connected layers and 3333 residual blocks. The input to the proposal network contains features of the current state, next state, and action. <|MaskedSetence|> Each residual block contains two dense layers. We use the skip-connection for the input and outp... | **A**: In each layer, we integrate the action with features from the previous layer, which amplifies the impact of actions.
**B**: The output of the proposal network is a 256256256256-dimensional vector, which contains the mean and standard deviation of the Gaussian latent variable 𝐳𝐳\mathbf{z}bold_z.
**C**: We fur... | ABC | ABC | CBA | ABC | Selection 1 |
The model has two parts. <|MaskedSetence|> We do that by applying any of the above mentioned VAEs111In this exposition we use unspervised trained VAEs as our base models but the framework also works with GAN-based or FLOW-based DGMs, supervised, semi-supervised or unsupervised. In the Appendix we present such impleme... | **A**: We can view this as a style transfer task and use a technique from [adaIN] to achieve our goal..
**B**: where we significantly constrain the capacity of the learned representation and heavily regularize the model to produce independent factors.
**C**: First, we apply a DGM to learn only the disentangled part, ... | CBA | ACB | CBA | CBA | Selection 4 |
<|MaskedSetence|> ‘Window operator’ is a cube of 3x3, each containing elements of 0,i,1,-1,2, and 2. Each element (or cell) is inputted in the same way as three pin structural computing on the upper and lower surfaces. <|MaskedSetence|> The expressions and functions of ) are as follows. <|MaskedSetence|> i, Black bo... | **A**: 1110, NULL: Transmits light that enters the upper and lower sides.
**B**: Now, we will define ‘window operators’ to have the same connection as a 3-pin based structural computer using the reverse signal pair described earlier.
**C**: I will call it this because it is a basic unit that makes up an organization ... | BCA | BCA | BCA | BCA | Selection 1 |
Forward selection is a simple, greedy feature selection algorithm (Guyon \BBA Elisseeff, \APACyear2003). It is a so-called wrapper method, which means it can be used in combination with any learner (Guyon \BBA Elisseeff, \APACyear2003). <|MaskedSetence|> One then proceeds to sequentially add the next “best” feature at... | **A**: In order to impose nonnegativity of the coefficients, we will use a slightly modified procedure which we will call nonnegative forward selection (NNFS).
**B**: Here we consider forward selection based on the Akaike Information Criterion (AIC).
**C**: The basic strategy is to start with a model with no features... | BCA | CBA | CBA | CBA | Selection 2 |
<|MaskedSetence|> We differentiate them in this subsection. To tackle the problem of anomaly detection in high-dimensional data, subspace anomaly detection methods, like those in [23, 24, 25, 1], have been proposed to detect anomalies with proximity-based approach in a subspace containing a subset of variables. <|Mas... | **A**: When determining subspaces, many methods (e.g., [24, 26, 25]) utilize the correlation of variables, and some (e.g., [27, 28]) randomly select variables.
**B**: Some readers may wonder what the differences are between DepAD and subspace anomaly detection approaches since both use a subset of variables for anomal... | BAC | BAC | BAC | CBA | Selection 3 |
<|MaskedSetence|> in Abbasi-Yadkori et al. [2011]), which is in contrast to the use of an exploration bonus as seen in Faury et al. [2020], Filippi et al. [2010]. Optimistic parameter search provides a cleaner description of the learning strategy. <|MaskedSetence|> <|MaskedSetence|> | **A**:
CB-MNL enforces optimism via an optimistic parameter search (e.g.
**B**: [2010] for a short discussion)..
**C**: In non-linear reward models, both approaches may not follow similar trajectory but may have overlapping analysis styles (see Filippi et al.
| ACB | ACB | ACB | ACB | Selection 1 |
Cross-scale graph network. <|MaskedSetence|> <|MaskedSetence|> Its architecture is illustrated in Fig. 4. The temporal branch contains a Conv1d(3,1)Conv1d31\textrm{Conv1d}(3,1)Conv1d ( 3 , 1 )222For conciseness, we use Conv1d(m,n)Conv1d𝑚𝑛\textrm{Conv1d}(m,n)Conv1d ( italic_m , italic_n ) to represent 1-D convolut... | **A**: The xGN module contains a temporal branch to aggregate features in a temporal neighborhood, and a graph branch to aggregate features from intra-scale and cross-scale locations.
**B**: Then it pools the aggregated features into a smaller temporal scale.
**C**: layer.
| ABC | ABC | ABC | ABC | Selection 1 |
Another open issue is the avoidance of hyperparameter tuning per se, as noted by E3. <|MaskedSetence|> <|MaskedSetence|> E1 expressed his interest in checking combinations of evolutionary optimization with the crossover and mutation process applied to the best-performing models (e.g., [YRK∗15]). <|MaskedSetence|> We... | **A**: Also, E3 stated that we could allow the user to specify the hyperparameters range at every stage and test alternative mutation strategies [CK05].
**B**: However, as the user usually adds—as few as possible—models to the ensembles, the hyperparameters’ evolution for the excluded algorithms will be infeasible.
*... | CAB | CAB | CAB | CAB | Selection 3 |
We presented a novel formulation for the isometric multi-shape matching problem. Our main idea is to simultaneously solve for shape-to-universe matchings and shape-to-universe functional maps. <|MaskedSetence|> <|MaskedSetence|> Our algorithm is efficient, straightforward to implement, and montonically increases the... | **A**: This contrasts the recent ConsistentZoomOut [31] method, which does not obtain cycle-consistent multi-matchings.
**B**: By doing so, we generalise the popular functional map framework to multi-matching, while guaranteeing cycle consistency, both for the shape-to-universe matchings, as well as for the shape-to-u... | BAC | CBA | BAC | BAC | Selection 1 |
<|MaskedSetence|> <|MaskedSetence|> In a few words, an antipodality graph has as vertex set some subgraph of G𝐺Gitalic_G, and two vertices are connected if the corresponding subgraphs of G𝐺Gitalic_G are antipodal. Unfortunately, we cannot build all the antipodality graphs by brute force because checking all possibl... | **A**:
The recognition algorithm RecognizePG for path graph is mainly built on path graphs’ characterization in [1].
**B**: This is done in Step 4, Step 5, and Step 6 that are the core of algorithm RecognizePG..
**C**: This characterization decomposes the input graph G𝐺Gitalic_G by clique separators as in [18], the... | BAC | ACB | ACB | ACB | Selection 4 |
The stochastic blockmodel (SBM) (SBM, ) is one of the most used models for community detection in which all nodes in the same community are assumed to have equal expected degrees. Some recent developments of SBM can be found in (abbe2017community, ) and references therein. Since in empirical network data sets, the deg... | **A**: However, in MMSB, nodes in the same communities still share the same degrees.
**B**: In this paper, we design community detection algorithms based on the DCMM model..
**C**: MMSB constructed a mixed membership stochastic blockmodel (MMSB) which is an extension of SBM by letting each node have different weight... | CBA | CAB | CAB | CAB | Selection 4 |
See, e.g., Welling and Teh (2011); Chen et al. (2014); Ma et al. (2015); Chen et al. (2015); Dubey et al. <|MaskedSetence|> <|MaskedSetence|> (2016); Dalalyan (2017); Chen et al. (2017); Raginsky et al. (2017); Brosse et al. (2018); Xu et al. (2018); Cheng and Bartlett (2018); Chatterji et al. (2018); Wibisono (2018)... | **A**: (2019); Wibisono (2019) and the references therein.
Among these works,.
**B**: (2016); Chen et al.
**C**: (2016); Vollmer et al.
| CBA | CBA | BAC | CBA | Selection 2 |
<|MaskedSetence|> It shows that the models learned by the regular RL algorithms indeed rely on the training scenario. MetaLight is more robust to various scenarios than Individual RL and PressLight, and it indicates the advantage of the meta-learning framework. <|MaskedSetence|> Overall, MetaVIM achieves the state-of... | **A**: The main reason is that: the task-specific information is modeled by the latent variable in our method, and the learned policy function could be adaptive to diverse latent variables.
**B**: The meta-learning framework could help to learn task-shared model.
**C**: 2) The performances of Individual RL and PressL... | CBA | CBA | BCA | CBA | Selection 2 |
Online bin packing has a long history of study. The simplest algorithm is NextFit, which places an item into its single open bin when possible; otherwise, it closes the bin (does not use it anymore) and opens a new bin for the item. FirstFit is another simple heuristic that places an item into the first bin of suffici... | **A**: Improving upon this performance requires more sophisticated algorithms, and many have been proposed in the literature..
**B**: BestFit works similarly, except that it places the item into the bin of minimum available capacity, which can still fit the item.
**C**: NextFit has a competitive ratio of 2, while bot... | BCA | ACB | BCA | BCA | Selection 4 |
To address the problem mentioned above, most of the methods extend the Chamfer loss function of basic AtlasNet with additional terms. Bednarik et al. (2020) added terms to prevent patch collapse, reduce patch overlap and calculate the exact surface properties analytically rather than approximating them. <|MaskedSetenc... | **A**: Another term enforces better spatial configuration of the mappings by minimizing a stitching error.
.
**B**: (2020b) introduced two additional terms to increase global consistency of the local mappings explicitly.
**C**: Deng et al.
| CBA | CBA | CBA | CBA | Selection 2 |
Paper organization. <|MaskedSetence|> <|MaskedSetence|> <|MaskedSetence|> In Section 4, we present the lower complexity bounds for saddle point problems without individual variables. Finally in Section 5, we show how the proposed algorithm can be applied to the problem computing Wasserstein barycenters .
. | **A**: In Section 3, we provide the main algorithm of the paper to solve such kind of problems.
**B**: Section 2 presents a saddle point problem of interest along with its decentralized reformulation.
**C**: This paper is organized as follows.
| ABC | CBA | CBA | CBA | Selection 3 |
<|MaskedSetence|> The minimum cycle basis (MCB) problem is the problem of finding a cycle basis such that the sum of the lengths (or edge weights) of its cycles is minimum. This problem was formulated by Stepanec [7] and Zykov [8] for general graphs and by Hubicka and Syslo [9] in the strictly fundamental class contex... | **A**: In [5] a unified perspective of the problem is presented.
**B**: Some applications of the MCB problem are described in [5, 11, 10, 12]..
**C**:
The length of a cycle is its number of edges.
| ACB | CAB | CAB | CAB | Selection 2 |
Feature selection is about choosing a subset of features from the pool of features available by that time. Feature selection methods can be generally divided into four high-level categories: (1) filter methods, (2) wrapper methods, (3) embedded methods, and (4) hybrid methods [16, 17, 18]. Our feature selection strate... | **A**: Also, instead of appending features progressively (called forward selection) or considering all features and then discarding some (known as backward elimination), we choose a stepwise selection approach.
**B**: Therefore, we start with all features, but we can add or remove any number of features at different s... | ABC | ABC | CAB | ABC | Selection 1 |
<|MaskedSetence|> High-precision trajectories or set points can be generated prior to the actual machining process following various optimization methods, including MPC, feed-forward PID control strategies, or iterative-learning control [6, 7], where friction or vibration-induced disturbances can be corrected. <|Mask... | **A**: In MPC, closed-loop performance is pushed to the limits only if the plant under control is accurately modeled, alternatively, the performance degrades due to imposed robustness constraints.
**B**: Instead of adapting the controller for the worst case scenarios, the prediction model can be selected to provide th... | ACB | CAB | CAB | CAB | Selection 3 |
We use the GQA visual question answering dataset [33] to highlight the challenges of using bias mitigation methods on real-world tasks. It has multiple sources of biases including imbalances in answer distribution, visual concept co-occurrences, question word correlations, and question type/answer distribution. <|Mask... | **A**: It is unclear how the explicit bias variables should be defined so that the methods can generalize to all minority groups.
**B**: It is unknown if bias mitigation methods can scale to hundreds and thousands of groups in GQA, yet natural tasks require such an ability.
.
**C**: GQA-OOD [36] divides the evaluatio... | ACB | ACB | ACB | ACB | Selection 3 |
<|MaskedSetence|> <|MaskedSetence|> They use online learning to fine-tune their model with the calibration samples.
Some studies investigate the relation between the gaze points and the saliency maps [125, 126]. Chang et al. utilize saliency information to adapt the gaze cestimation algorithm to a new user without ex... | **A**: Salvalaio et al. implicitly collect calibration data when users are using computers.
**B**: They collect data when the user is clicking a mouse, this is based on the assumption that users are gazing at the position of the cursor when clicking the mouse [146].
**C**: They minimize the difference between the pro... | ABC | BCA | ABC | ABC | Selection 3 |
Experimental results are carried out on Real-world Masked Face Recognition Dataset (RMFRD) and Simulated Masked Face Recognition Dataset (SMFRD) presented in wang2020masked . <|MaskedSetence|> To do so, we apply a cropping filter in order to obtain only the informative regions of the masked face (i.e. <|MaskedSetence... | **A**: This strategy, however, is a difficult and highly time-consuming process.
.
**B**: forehead and eyes).
**C**: We start by localizing the mask region.
| CBA | BAC | CBA | CBA | Selection 1 |
Sized types are a type-oriented formulation of size-change termination [LJBA01] for rewrite systems [TG03, BR09]. Sized (co)inductive types [BFG+04, Bla04, Abe08, AP16] gave way to sized mixed inductive-coinductive types [Abe12, AP16]. <|MaskedSetence|> <|MaskedSetence|> As we mentioned in the introduction, we use un... | **A**: However, the state of the art [Abe12, AP16, CLB23] supports polymorphic, higher-kinded, and dependent types, which we aim to incorporate in future work.
.
**B**: We present, to our knowledge, the first sized type system for a concurrent programming language as well as the first system to combine both features f... | CBA | CBA | CBA | CAB | Selection 2 |
<|MaskedSetence|> First, suppose an owner rents the cloud’s resources for media sharing, the owner and the cloud execute Part 1 as shown in Fig. 5. Then, suppose the k𝑘kitalic_k-th user makes a request indicating that he/she wants to access one of the owner’s media content 𝐦𝐦\mathbf{m}bold_m, the entities execute P... | **A**: Finally, the arbitration and traitor tracing process follows the same approach of FairCMS-I and is thus omitted here.
.
**B**: The whole FairCMS-II scheme is summarized as follows.
**C**: 6.
| BCA | BCA | CBA | BCA | Selection 1 |
Currently, Graph Neural Networks (GNN) Kipf and Welling (2017); Hamilton et al. <|MaskedSetence|> (2018) have recently emerged as an effective class of models for capturing high-order relationships between nodes in a graph and have achieved state-of-the-art results on a variety of tasks such as computer vision Li et ... | **A**: (2017), neural language processing Marcheggiani and Titov (2017); Yao et al.
**B**: (2019)..
**C**: (2017); Veličković et al.
| CAB | CAB | CAB | ACB | Selection 1 |
Self-concordant functions have received strong interest in recent years due to the attractive properties that they allow to prove for many statistical estimation settings [Marteau-Ferey et al., 2019, Ostrovskii & Bach, 2021]. <|MaskedSetence|> For example, the logistic loss function used in logistic regression is not ... | **A**: The original definition of self-concordance has been expanded and generalized since its inception, as many objective functions of interest have self-concordant-like properties without satisfying the strict definition of self-concordance.
**B**: [2015], in which more general properties of these
pseudo-self-conco... | ABC | ABC | ABC | BCA | Selection 1 |
Our algorithm executes several methods (invoked within the loop starting at Algorithm 2 of Algorithm 2), and for most of them it makes a fresh pass over the edges. <|MaskedSetence|> Precisely, the routines are: (1) extend structures along active paths (Extend-Active-Paths), (2) check for edge augmentations (Check-for... | **A**: The term Pass-Bundle refers to multiple passes during which those routines are executed.
**B**: In total, a Pass-Bundle requires 3333 passes..
**C**: The Backtrack-Stuck-Structures method backtracks active paths that were not extended, but does not require a fresh pass.
| CBA | ACB | ACB | ACB | Selection 2 |
Setting. <|MaskedSetence|> This is one of the default learning settings. Based on these settings, we build our settings using the intuition of algorithms (for details about tuning and intuition of our Algorithms, see Section 5.2). <|MaskedSetence|> That is why we need carefully choose T𝑇Titalic_T (the number of inn... | **A**: To train ResNet18 in CIFAR-10, one can use stochastic gradient descent with momentum 0.90.90.90.9, the learning rate of 0.10.10.10.1 and a batch size of 128128128128 (40404040 batches = 1111 epoch).
**B**: In order for the comparison of Algorithm 1 and Algorithm 3 to be fair, it is necessary to balance two thin... | BAC | ABC | ABC | ABC | Selection 4 |
<|MaskedSetence|> <|MaskedSetence|> The component that determines the distribution of policies that the oracle responds to is called the meta-solver (MS). The MS operates on the meta-game (MG), which is a payoff tensor estimated by measuring the expected return (ER) of policies against one another. This is a NF game,... | **A**:
PSRO consists of a response oracle that estimates the best response (BR) to a joint distribution of policies.
**B**: The set of deterministic policies can be huge and that of stochastic policies is infinite, therefore PSRO only considers a subset of game policies: the ones found by the BR over all iterations s... | BCA | ACB | ACB | ACB | Selection 4 |
<|MaskedSetence|> The following lemma gives a useful and intuitive characterization of the quantity that the Bayes stability definition requires be bounded. Simply put, the Bayes factor K(⋅,⋅)𝐾⋅⋅{K}\left(\cdot,\cdot\right)italic_K ( ⋅ , ⋅ ) (defined in the lemma below) represents the amount of information leaked abo... | **A**: Its corresponding version for arbitrary queries are presented in Section C.2..
**B**: Since achieving posterior accuracy is relatively straightforward, guaranteeing Bayes stability is the main challenge in leveraging this theorem to achieve distribution accuracy with respect to adaptively chosen queries.
**C**... | BAC | BCA | BCA | BCA | Selection 2 |
The remainder of the paper is organized as follows. After presenting preliminaries on graphs and sets in Section 2, we prove the mentioned hardness results in Section 3. <|MaskedSetence|> In Section 5 we show how color coding can be used to find a large feedback vertex cut, if one exists. <|MaskedSetence|> Our main r... | **A**: We present structural properties of antlers and how they combine in Section 4.
**B**:
.
**C**: We also prove that, given a large feedback vertex cut, we can shrink it while preserving the antlers in the graph.
| ACB | ACB | ACB | ABC | Selection 3 |
<|MaskedSetence|> Note that both artistic style transfer [37, 56, 118] and photorealistic style transfer [103, 82] belong to style transfer. <|MaskedSetence|> There are two main differences between image harmonization and photorealistic style transfer. 1) Firstly, image harmonization adjusts the foreground appearance... | **A**: 2) Secondly, the definition of “style” in photorealistic style transfer is unclear and coarsely depends on the employed style loss (e.g., Gram matrix loss [37], AdaIn loss [56]).
**B**:
Image harmonization is closely related to style transfer.
**C**: Image harmonization is closer to photorealistic style trans... | BCA | BCA | BCA | CAB | Selection 3 |
In brief, the creation and implementation of a comprehensive urban dataset encounter two major challenges. Firstly, urban data are usually fragmented across different entities, such as governmental bodies and private enterprises, resulting in disparities in data acquisition and processing protocols. These differences ... | **A**: Secondly, beyond data collection, identifying interdependencies among various datasets is critical to enhance performance by sharing and transferring relevant knowledge.
**B**: Consequently, integrating these disparate data sources into a standardized format with aligned range for research purposes poses a sign... | BAC | BAC | BAC | ACB | Selection 1 |
Although a variety of methods was considered, it is not feasible to include all of them. <|MaskedSetence|> The main reason for this omission is the large number of choices in terms of priors and approximations, both of which strongly depend on the problem at hand. <|MaskedSetence|> For general regression models the ... | **A**: On the level of calibration there are also some methods that were not included in this paper, mostly because they were either too specific or too complex for simple regression problems.
**B**: The most important omission is a more detailed overview of Bayesian neural networks (although one can argue, as was don... | CBA | BAC | BAC | BAC | Selection 2 |
<|MaskedSetence|> The former marks the beginning of a new bar, while the latter points to a discrete position within a bar. <|MaskedSetence|> <|MaskedSetence|> As depicted in Fig. 1(a), we use a Sub-bar token before each musical note, which comprises two consecutive tokens of Pitch and Duration. In other words, the ... | **A**: The REMI representation \parencitehuang2020pop for MIDI performances uses Bar and Sub-bar tokens to represent the advancement in time.
**B**: We prefer our naming for it is musically more accurate—our Sub-bar tokens are subdivisions of a bar (i.e., dividing a bar into 16 points), not subdivisions of a beat (i.e... | ACB | ACB | ACB | ACB | Selection 3 |
Recently, there are also investigations on semantic communications for other transmission contents, such as image and speech. <|MaskedSetence|> <|MaskedSetence|> Similar to text transmission, IoT applications for image transmission have been carried out. <|MaskedSetence|> A deep joint source-channel coding architec... | **A**: Based on JSCC, an image transmission system, integrating channel output feedback, can improve image reconstruction[15].
**B**: Particularly, a joint image transmission-recognition system has been developed in[16] to achieve high recognition accuracy.
**C**: A DL-enabled semantic communication system for image ... | CAB | CAB | BCA | CAB | Selection 2 |
<|MaskedSetence|> This ensures that supervision is densely transmitted from labeled to unlabeled points within a given sample. Given the potential for non-overlapping classes between input pairs, which could introduce noise during the supervision propagation, our strategy involves training the network using the CSFR m... | **A**: They are redundant for generating the final segmentation predictions during the inference phase.
**B**:
In the subsequent training stage, our ISFR module facilitates the transmission of supervision from labeled to unlabeled points within each individual sample, once again utilizing feature reallocation based o... | BCA | BAC | BAC | BAC | Selection 2 |
<|MaskedSetence|> The KITTI dataset [11] provides widely used benchmarks for various visual tasks in the autonomous driving, including 2D Object detection, Average Orientation Similarity (AOS), Bird’s Eye View (BEV), and 3D Object Detection. The official data set contains 7481 training and 7518 test images with 2D and... | **A**: Each class uses different IoU standards for further evaluations.
**B**: We report the average accuracy (APAP\rm{AP}roman_AP) for each task under three different settings: easy, moderate, and hard, as defined in [11].
**C**: Setup.
| ABC | CBA | CBA | CBA | Selection 3 |
<|MaskedSetence|> Its ground truth is annotated with word-level quadrangles. <|MaskedSetence|> <|MaskedSetence|> It contains 300 training images and 200 testing images with word-level annotation. Here, we follow the previous methods [35, 8] and add 400 training images from TR400 [46] to extend this dataset.. | **A**: ICDAR2015 [44] includes multi-orientated and small-scale text instances.
**B**:
MSRA-TD500 [45] is dedicated to detecting multi-oriented long non-Latin texts.
**C**: It contains 1,000 training and 500 testing images.
| ACB | ACB | BCA | ACB | Selection 1 |
Assume that each memory address of a 64-bit system can be stored in an element 8 bytes in size, and the number of occurrences of an individual IP address is no more than 264superscript2642^{64}2 start_POSTSUPERSCRIPT 64 end_POSTSUPERSCRIPT. <|MaskedSetence|> The array is regarded as a memory block that contains a con... | **A**: There is an array of size 256 elements that consists of 256×82568256\times 8256 × 8 bytes of memory.
**B**: 1.
**C**: Therefore, the position of the array can be indexed by the integer value of a particular part of the IP address.
| ABC | ABC | ABC | ABC | Selection 4 |
The outline of the remainder of this paper is as follows. <|MaskedSetence|> <|MaskedSetence|> Some additive Schur complement based preconditioners are constructed and the corresponding known results in the literature are recalled in Section 4 for twofold saddle point problems. Generalizations to n𝑛nitalic_n-tuple ca... | **A**: Furthermore, we extend these results to the n𝑛nitalic_n-tuple saddle point problem in Section 3.
**B**: In section 2, we briefly recall the classic saddle point problem and its Schur complement, and introduce the twofold saddle point problem and the form of Schur complement, we then construct and analyze the b... | CAB | BAC | BAC | BAC | Selection 2 |
However, in cases when the labels are sensitive and sharing the labels for a sample ID across silos is not feasible, the label information for a sample ID may only be present in a client in one silo. In this case, we could modify our algorithm in the following way, similar to (Liu et al., 2020a): the clients in all sil... | **A**: This modification would significantly increase the communication cost of the algorithm.
**B**: We note that the modified algorithm is mathematically equivalent to TDCD, albeit with a higher communication cost.
**C**: Hence, the convergence analysis given in Section 4 can be trivially extended to this case.
.
| ABC | ABC | ABC | ACB | Selection 1 |
<|MaskedSetence|> <|MaskedSetence|> The properties of pseudospectra are also discussed, along with a characterization of the pseudospectra for normal matrices. Additionally, for diagonalizable but not necessarily normal matrices, the corresponding Bauer-Fike theorem is presented, which can be found in (trefethen2005s... | **A**: In the book, four different definitions of matrix pseudospectra are introduced and shown to be equivalent under certain conditions.
**B**: The pseudospectra of finite-dimensional matrices and their extension to linear operators in Banach space have been extensively investigated and summarized in the classical b... | BAC | BAC | BAC | BAC | Selection 2 |
<|MaskedSetence|> To further highlight the two-stream dual generation architecture, we compare it with a multi-task single-stream network, which is tailed by two branches to model the image structure and texture simultaneously. We enlarge its channels to make it have the same amount of parameters as the proposed netwo... | **A**: Quantitative results in Table 2 also validate the advantages of texture and structure dual generation.
.
**B**: As shown in Figure 7 (c), the two-stream architecture exhibits superior performance with more visually reasonable structures and detailed textures.
**C**: On Two-stream Network Architecture.
| CBA | CBA | CBA | ACB | Selection 2 |
<|MaskedSetence|> <|MaskedSetence|> We used expert data to generate labels for supervised training. When offline datasets are available, which is the case for the environments considered in this paper111The dataset for INT or Sokoban can be easily generated or are publicly available. For the Rubik’s Cube, we use rand... | **A**: The main advantages of this subgoal objective are simplicity and empirical efficiency.
**B**: Consequently, this method is often taken when dealing with complex domains (see e.g.
**C**: We train the transformer with the objective to predict the k𝑘kitalic_k-th step ahead.
| CAB | BCA | CAB | CAB | Selection 3 |
Ablation study is thus made to investigate how glyph and phonetic features bring improvement by themselves. We separately add glyph embedding or phonetic embedding to pre-trained language models for comparisons. <|MaskedSetence|> <|MaskedSetence|> On one hand, named entities in different datasets may rely on one of o... | **A**: From the results on all four datasets, it is clear that models with the glyph or phonetic embedding almost all perform better than models with pure semantic embedding, which means that extra patterns from glyph and phonetic domains are all helpful in the NER task, strengthening the original model ability.
Howeve... | CBA | ACB | ACB | ACB | Selection 2 |
<|MaskedSetence|> The shared modules learn shared features from multiple tasks. Since the shared modules can learn from many tasks, they can be sufficiently trained and can generalize better, which is particularly meaningful for low-resource scenarios. On the other hand, task-specific modules learn features that are s... | **A**: The idea behind the modular MTL architecture is simple: breaking an MTL model into shared modules and task-specific modules.
**B**: Compared with shared modules, task-specific modules are usually much smaller and thus less likely to suffer from overfitting caused by insufficient training data.
**C**: The robus... | BCA | ABC | ABC | ABC | Selection 2 |
The templates are intended to approximate the final look and page length of the articles/papers. <|MaskedSetence|> <|MaskedSetence|> The structure of the LaTeXfiles, as designed, enable easy conversion to XML for the composition systems used by the IEEE’s outsource vendors. <|MaskedSetence|> Have you looked at your ... | **A**: The XML files are used to produce the final print/IEEEXplore® pdf and then converted to HTML for IEEEXplore®.
**B**: Therefore, they are NOT intended to be the final produced work that is displayed in print or on IEEEXplore®.
**C**: They will help to give the authors an approximation of the number of pages tha... | BCA | BCA | BCA | BCA | Selection 1 |
<|MaskedSetence|> <|MaskedSetence|> The full set of survey questions is reproduced in Appendix LABEL:app:questions. In order to distill answers from these questions into a reasonable number of attributes to describe reciprocity and trust, we performed a principal components analysis and selected the top two principal... | **A**: The first principal component, referred to as overall reciprocity, captures both positive and negative reciprocity, while the second principal component, referred to as positive reciprocity, favors positive reciprocity and eschews negative reciprocity..
**B**: All of these questions were aimed at eliciting a su... | CBA | CBA | CAB | CBA | Selection 2 |
Unsupervised Learning: The simulated paired images have poor versatility, while the real paired images are difficult to collect. To address this issue, some methods began to try to no longer use paired LR-HR images for training. We often call this type of method an unsupervised learning method. <|MaskedSetence|> <|Ma... | **A**: Among them, some researchers first learn the HR-to-LR degradation and use it to construct datasets for training the model, while other researchers design cycle-in-cycle models to learn the LR-to-HR and HR-to-LR mappings simultaneously.
**B**: This type of unsupervised method no longer uses paired LR-HR images f... | BCA | BCA | ACB | BCA | Selection 4 |
There have been some works where coordinate-based networks are used as a core for a generative model using techniques such as a hypernetwork predicting the weights of a sample coordinate [11], or by modulating the weights of a base coordinate [12]. <|MaskedSetence|> Finally, Local Implicit Image Functions introduce... | **A**: These approaches are fundamentally different as they attempt to create a wide generative model based on a large-scale dataset, while our approach focuses on data-agnostic internal learning tasks and uses a disparate architecture.
**B**: Additionally, internal learning approaches relying on the priors contained ... | ABC | ACB | ACB | ACB | Selection 2 |
<|MaskedSetence|> <|MaskedSetence|> Chow,, 1957; Sayedi et al.,, 2010; Wiener and El-Yaniv,, 2011) the learner may decline to label items, thus mitigating the risk of labelling when they have high uncertainty. Conversely, in classification with selective sampling (Cesa-Bianchi et al.,, 2009; Orabona and Cesa-Bianchi,... | **A**: The apple tasting problem is not the only variant of online classification where labels are not revealed in every round.
**B**: In selective classification (or classification with a reject (or abstention) option) (e.g.
**C**: Both of these variants differ from apple tasting in that they have a more complex act... | BAC | ABC | ABC | ABC | Selection 2 |
The introduction of smart sampling strategies allows scaling to larger memories. <|MaskedSetence|> <|MaskedSetence|> In our experiments, this led priority-based strategies to learn a quasi-uniform priority distribution. <|MaskedSetence|> For instance, a parametric model could be trained to embed input texts near th... | **A**: This suggests a possible direction for improving the adoption of input-conditioned sampling strategies compatible with SS regularization.
**B**: However, the introduced priority-based strategies lack proper conditioning on the input, but rather learn the dataset-level importance of each memory slot.
**C**: Thi... | BCA | BCA | CBA | BCA | Selection 2 |
<|MaskedSetence|> (2019); Zhou et al. <|MaskedSetence|> (2021); Li et al. <|MaskedSetence|> (2021), has contributed to the improvement of coherent sentiment learning.
These studies explored the effectiveness of syntax information in ABSC, which mitigates issues related to sentiment coherency extraction.. | **A**: However, the progress of sentiment dependency-based methods, such as the work by Zhang et al.
**B**: (2021a); Dai et al.
**C**: (2020); Tian et al.
| BAC | ACB | ACB | ACB | Selection 3 |
QuantumNAT comprises a three-stage pipeline. The first step, post-measurement normalization normalizes the measurement outcomes on each quantum bit (qubit) across data samples, thus removing the quantum error-induced distribution shift. <|MaskedSetence|> <|MaskedSetence|> <|MaskedSetence|> Finally, post-measurement ... | **A**: The error gate types and probabilities are obtained from hardware-specific realistic quantum noise models provided by QC vendors.
**B**: During training, we iteratively sample error gates, insert them to PQC, and updates weights.
**C**: Furthermore, we inject noise to the PQC training process by performing err... | CAB | CAB | CAB | ACB | Selection 1 |
In this section, we describe all the components of the proposed EDA approach. The pipeline of the proposed EDA is illustrated in Fig. 2. First, we describe the sequential retinal events, and introduce an asynchronous fusion phase to gather the sequential retinal events, as illustrated in Fig. 2(a)-(d). Then, we present... | **A**: 2(g)-(h).
.
**B**: Finally, we give the details of the model hypothesis selection, to robustly perform the final data association, as illustrated in Fig.
**C**: 2(e)-(f).
| CAB | CBA | CBA | CBA | Selection 2 |
<|MaskedSetence|> <|MaskedSetence|> Recently, contrastive learning (CL) [38, 39, 40, 41], which discriminates positive pairs against negative pairs, achieved state-of-the-art performance in various vision tasks. <|MaskedSetence|> To fully utilize negative samples, [45, 46, 47, 48] explore hard samples in the momentu... | **A**:
Early SSL methods design hand-crafted pretext tasks [30, 31, 32, 33], which rely on somewhat ad-hoc heuristics and have limited abilities to capture practically useful representations.
**B**: Another popular form is clustering-based methods [34, 35, 36, 37] learning discriminative representation by offline or ... | ABC | ABC | ABC | ABC | Selection 1 |
<|MaskedSetence|> We use a population size of 100. We randomly sample 100 sub-networks satisfying the constraints to form the first generation of population. <|MaskedSetence|> Then we perform crossover to generate 50 new candidates and mutation to generate another 50, forming a new generation. The mutation rate is 0.... | **A**: We used evolutionary search to find the best sub-network architecture under certain constraints.
**B**: We repeat the process for 30 iterations and choose the sub-network with the highest accuracy on the split validation set.
.
**C**: For each iteration, we only keep the top-20 candidates with the highest accu... | ACB | ACB | ACB | ACB | Selection 2 |
<|MaskedSetence|> The asymmetry lies in the differences of GNN-based encoders’ message-passing schemes. Besides, graph encoders in CGCL are supposed to be complementary for a stronger fitting ability. Specifically, high complementarity indicate that encoders together carry less redundant parameters. <|MaskedSetence|>... | **A**: Compared with the state-of-the-art methods, CGCL demonstrates better generalization on various datasets and achieves better results without using extra handcrafted data augmentations.
**B**: To cope with the problem of model collapse, we devise the asymmetric structure for CGCL.
**C**: For a further theoretica... | BCA | CAB | BCA | BCA | Selection 4 |
The work of Piotr Miłoś was supported by the Polish National Science Center grant UMO-2017/26/E/ST6/00622. The work of Tomasz Korbak was supported by the Leverhulme Doctoral Scholarship. We gratefully acknowledge Polish high-performance computing infrastructure PLGrid (HPC Centers: ACK Cyfronet AGH, PCSS) for providin... | **A**: Our experiments were managed using https://neptune.ai.
**B**: PLG/2019/012498.
**C**: We would like to thank the Neptune team for providing us access to the team version and technical support..
| BAC | BAC | BAC | BAC | Selection 2 |
<|MaskedSetence|> Indeed, the lack of systematic methods to construct valid CBFs is a main bottleneck. <|MaskedSetence|> <|MaskedSetence|> Finding CBFs poses additional challenges in terms of the control input resulting in bilinear SOS programming as presented in [32, 33] and summarized in [34]. The work in [35] con... | **A**: For certain types of mechanical systems under input constraints, analytic CBFs can be constructed [30].
**B**: Learning CBFs: An open problem is how valid CBFs can be constructed.
**C**: The construction of polynomial barrier functions towards certifying safety for polynomial systems by using sum-of-squares (... | BAC | BAC | BAC | ABC | Selection 2 |
In CoauthorshipsNet, node means scientist and weights mean coauthorship, where weights are assigned by the original papers. <|MaskedSetence|> The CoauthorshipsNet has 1589 nodes, however its adjacency matrix is disconnected. <|MaskedSetence|> For convenience, we use CoauthorshipsNet1589 to denote the original network... | **A**: Among the 1589 nodes, there are totally 396 disconnected components, and only 379 nodes fall in the largest connected component.
**B**: For this network, there is no ground truth about nodes labels, and the numbers of communities are unknown.
**C**: To find the number of communities for CoauthorshipsNet, we pl... | BAC | BCA | BAC | BAC | Selection 3 |
End of preview. Expand in Data Studio
README.md exists but content is empty.
- Downloads last month
- 9