The authors would like to thank Kunal Talwar for suggesting the study of this problem, and Adam Klivans, Frederic Koehler, Ankur Moitra, and Shanshan Wu for helpful and inspiring conversations. {CJK*}UTF8gbsn GK would like to thank Chengdu Style Res…

In this paper, we investigate the connection between robustness and simplicity and show an inverse relationship, i.e., it is increasingly difficult to perturb an example when decreasing the complexity of the classification task. Motivated by this ob…

In this paper, we described a new approach, DeepWeiSurv, to the survival analysis. The key role of DeepWeiSurv is to predict the parameters of a mixture of Weibull distributions with respect to the covariates in presence of right-censored data. In a…

We explored the practical problem of enforcing group-based fairness for binary classification given noisy protected group information. In addition to providing novel theoretical analysis of the naïve approach of only enforcing fairness on the noisy …

This work is supported by the ERC under the European Union’s Horizon 2020 Research and Innovation Program 714608-SMiLe, as well as by the French Agence Nationale de la Recherche under grant ANR-17-CE23-0023-01 PAIL and ANR-19-P3IA-0001 PRAIRIE. We a…

We presented new results on the algorithm of variable elimination that exploit functional dependencies using dense factors, allowing one to benefit from tensor-based technologies for more efficient inference and learning. We also presented case stud…

Gaussian processes are the gold standard for many real-world modeling problems, especially in cases where a model's success hinges upon its ability to faithfully represent predictive uncertainty. These problems typically exist as parts of larger fra…

In this paper we present Transformer Hawkes Process, a framework for analyzing event streams. Event sequence data are common in our daily life, and they exhibit sophisticated short-term and long-term dependencies. Our proposed model utilizes the sel…

In this paper, we proposed a new curriculum-style method for image inpainting by progressively growing the mask regions. Experiments show that our method generates realistic and plausible images, even with large masked regions. Further, it improves …

In this work, we introduced a p-value to measure Knockoff importance and designed a knockoffs bootstrapping scheme that leverages this quantity. With this we are able to tame the instability inherent to the original Knockoff procedure. Analysis show…

We first explore some basic properties of J(DP,T), which help a lot in convergence analysis of Algorithm 1. We have already shown that J(DP,T) are rational function of each entry in T. As for the DP part, it is indeed partially convex, which is give…

imagine is a learning architecture enabling autonomous learning by leveraging NL interactions with a social partner. As other algorithms from the imgep family, imagine sets its own goals and builds behavioral repertoires without external rewards. Th…

In this paper we have presented a machine learning model which changes the features during the training in favour of finding a more useful set of features. We tested our method on image classification and linear regression tasks. With the help of ou…

This paper presents a comprehensive analysis and comparison of the existing models for aligning embedding spaces w.r.t. the research questions mentioned in Sect. 2. This section summarizes the lessons learnt from the comparison and analysis of diffe…

Throughout this paper we have studied sparsity-induced overfitting using novel visualizations. We conclude that sparse layer responses can be encoded by the entropy of the receptive field activation vector. When overfitted, filters have been trained…

This work contains an exhaustive set of statistical procedures and examples of their use in the comparison of results from experimental studies of optimisation algorithms, specifically in the scenario of a competition. In this paper we have describe…