The pervasive brittleness of deep neural networks [Sze 14; Eng 19; HD19; Ath 18] has attracted significant attention in recent years. Particularly worrisome is the phenomenon of adversarial examples [Big 13; Sze 14], imperceptibly perturbed natural inputs that induce erroneous predictions in state-of-the-art classifiers. Previous work has proposed a variety of explanations for this phenomenon, ranging from theoretical models [Sch 18; BPR18] to arguments based on concentration of measure in high-dimensions [Gil 18; MDM18; Sha 19a]. These theories, however, are often unable to fully capture behaviors we observe in practice (we discuss this further in Section 5). More broadly, previous work in the field tends to view adversarial examples as aberrations arising either from the high dimensional nature of the input space or statistical fluctuations in the training data [Sze 14; GSS15; Gil 18]. From this point of view, it is natural to treat adversarial robustness as a goal that can be disentangled and pursued independently from maximizing accuracy [Mad 18; SHS19; Sug 19], either through improved standard regularization methods [TG16] or pre/post-processing of network inputs/outputs [Ues 18; CW17a; He 17].

Recommending appropriate items to users is crucial in many e-commerce platforms that contain implicit data as users' browsing, purchasing and streaming history. One common approach consists in selecting the N most relevant items to each user, for a given N, which is called top-N recommendation. To do so, recommender systems rely on various kinds of information, like item and user features, past interest of users for items, browsing history and trust between users. However, they often use only one or two such pieces of information, which limits their performance. In this paper, we design and implement GraFC2T2, a general graph-based framework to easily combine and compare various kinds of side information for top-N recommendation. It encodes content-based features, temporal and trust information into a complex graph, and uses personalized PageRank on this graph to perform recommendation. We conduct experiments on Epinions and Ciao datasets, and compare obtained performances using F1-score, Hit ratio and MAP evaluation metrics, to systems based on matrix factorization and deep learning. This shows that our framework is convenient for such explorations, and that combining different kinds of information indeed improves recommendation in general.

Multi-domain image-to-image translation is a problem where the goal is to learn mappings among multiple domains. This problem is challenging in terms of scalability because it requires the learning of numerous mappings, the number of which increases proportional to the number of domains. However, generative adversarial networks (GANs) have emerged recently as a powerful framework for this problem. In particular, label-conditional extensions (e.g., StarGAN) have become a promising solution owing to their ability to address this problem using only a single unified model. Nonetheless, a limitation is that they rely on the availability of large-scale clean-labeled data, which are often laborious or impractical to collect in a real-world scenario. To overcome this limitation, we propose a novel model called the label-noise robust image-to-image translation model (RMIT) that can learn a clean label conditional generator even when noisy labeled data are only available. In particular, we propose a novel loss called the virtual cycle consistency loss that is able to regularize cyclic reconstruction independently of noisy labeled data, as well as we introduce advanced techniques to boost the performance in practice. Our experimental results demonstrate that RMIT is useful for obtaining label-noise robustness in various settings including synthetic and real-world noise.

There is a growing need for fast and accurate methods for testing developmental neurotoxicity across several chemical exposure sources. Current approaches, such as in vivo animal studies, and assays of animal and human primary cell cultures, suffer from challenges related to time, cost, and applicability to human physiology. We previously demonstrated success employing machine learning to predict developmental neurotoxicity using gene expression data collected from human 3D tissue models exposed to various compounds. The 3D model is biologically similar to developing neural structures, but its complexity necessitates extensive expertise and effort to employ. By instead focusing solely on constructing an assay of developmental neurotoxicity, we propose that a simpler 2D tissue model may prove sufficient. We thus compare the accuracy of predictive models trained on data from a 2D tissue model with those trained on data from a 3D tissue model, and find the 2D model to be substantially more accurate. Furthermore, we find the 2D model to be more robust under stringent gene set selection, whereas the 3D model suffers substantial accuracy degradation. While both approaches have advantages and disadvantages, we propose that our described 2D approach could be a valuable tool for decision makers when prioritizing neurotoxicity screening.

The Unbiased Learning-to-Rank framework has been recently proposed as a general approach to systematically remove biases, such as position bias, from learning-to-rank models. The method takes two steps - estimating click propensities and using them to train unbiased models. Most common methods proposed in the literature for estimating propensities involve some degree of intervention in the live search engine. An alternative approach proposed recently uses an Expectation Maximization (EM) algorithm to estimate propensities by using ranking features for estimating relevances. In this work we propose a novel method to directly estimate propensities which does not use any intervention in live search or rely on modeling relevance. Rather, we take advantage of the fact that the same query-document pair may naturally change ranks over time. This typically occurs for eCommerce search because of change of popularity of items over time, existence of time dependent ranking features, or addition or removal of items to the index (an item getting sold or a new item being listed). However, our method is general and can be applied to any search engine for which the rank of the same document may naturally change over time for the same query. We derive a simple likelihood function that depends on propensities only, and by maximizing the likelihood we are able to get estimates of the propensities. We apply this method to eBay search data to estimate click propensities for web and mobile search and compare these with estimates using the EM method. We also use simulated data to show that the method gives reliable estimates of the "true" simulated propensities. Finally, we train an unbiased learning-to-rank model for eBay search using the estimated propensities and show that it outperforms both baselines - one without position bias correction and one with position bias correction using the EM method.

In e-commerce, product content, especially product images have a significant influence on a customer's journey from product discovery to evaluation and finally, purchase decision. Since many e-commerce retailers sell items from other third-party marketplace sellers besides their own, the content published by both internal and external content creators needs to be monitored and enriched, wherever possible. Despite guidelines and warnings, product listings that contain offensive and non-compliant images continue to enter catalogs. Offensive and non-compliant content can include a wide range of objects, logos, and banners conveying violent, sexually explicit, racist, or promotional messages. Such images can severely damage the customer experience, lead to legal issues, and erode the company brand. In this paper, we present a machine learning driven offensive and non-compliant image detection system for extremely large e-commerce catalogs. This system proactively detects and removes such content before they are published to the customer-facing website. This paper delves into the unique challenges of applying machine learning to real-world data from retail domain with hundreds of millions of product images. We demonstrate how we resolve the issue of non-compliant content that appears across tens of thousands of product categories. We also describe how we deal with the sheer variety in which each single non-compliant scenario appears. This paper showcases a number of practical yet unique approaches such as representative training data creation that are critical to solve an extremely rarely occurring problem. In summary, our system combines state-of-the-art image classification and object detection techniques, and fine tunes them with internal data to develop a solution customized for a massive, diverse, and constantly evolving product catalog.

The Hearthstone AI framework and competition motivates the development of artificial intelligence agents that can play collectible card games. A special feature of those games is the high variety of cards, which can be chosen by the players to create their own decks. In contrast to simpler card games, the value of many cards is determined by their possible synergies. The vast amount of possible decks, the randomness of the game, as well as the restricted information during the player's turn offer quite a hard challenge for the development of game-playing agents. This short paper introduces the competition framework and goes into more detail on the problems and challenges that need to be faced during the development process.

In this paper, we present a deep learning-based (DL-based) algorithm, as a purely mathematical platform, for providing intuitive understanding of the properties of electromagnetic (EM) wave-matter interaction in nanostructures. This approach is based on using the dimensionality reduction (DR) technique to significantly reduce the dimensionality of a generic EM wave-matter interaction problem without imposing significant error. Such an approach implicitly provides useful information about the role of different features (or design parameters such as geometry) of the nanostructure in its response functionality. To demonstrate the practical capabilities of this DL-based technique, we apply it to a reconfigurable optical metadevice enabling dual-band and triple-band optical absorption in the telecommunication window. Combination of the proposed approach with existing commercialized full-wave simulation tools offers a powerful toolkit to extract basic mechanisms of wave-matter interaction in complex EM devices and facilitate the design and optimization of nanostructures for a large range of applications including imaging, spectroscopy, and signal processing. It is worth to mention that the demonstrated approach is general and can be used in a large range of problems as long as enough training data can be provided.

Tactical decision making for autonomous driving is challenging due to the diversity of environments, the uncertainty in the sensor information, and the complex interaction with other road users. This paper introduces a general framework for tactical decision making, which combines the concepts of planning and learning, in the form of Monte Carlo tree search and deep reinforcement learning. The method is based on the AlphaGo Zero algorithm, which is extended to a domain with a continuous state space where self-play cannot be used. The framework is applied to two different highway driving cases in a simulated environment and it is shown to perform better than a commonly used baseline method. The strength of combining planning and learning is also illustrated by a comparison to using the Monte Carlo tree search or the neural network policy separately.

Gene interaction graphs aim to capture various relationships between genes and can be used to create more biologically-intuitive models for machine learning. There are many such graphs available which can differ in the number of genes and edges covered. In this work, we attempt to evaluate the biases provided by those graphs through utilizing them for 'Single Gene Inference' (SGI) which serves as, what we believe is, a proxy for more relevant prediction tasks. The SGI task assesses how well a gene's neighbors in a particular graph can 'explain' the gene itself in comparison to the baseline of using all the genes in the dataset. We evaluate seven major gene interaction graphs created by different research groups on two distinct datasets, TCGA and GTEx. We find that some graphs perform on par with the unbiased baseline for most genes with a significantly smaller feature set.