Despite recent advances in architectures for mobile devices, deep learning computational requirements remains prohibitive for most embedded devices. To address that issue, we envision sharing the computational costs of inference between local devices and the cloud, taking advantage of the compression performed by the first layers of the networks to reduce communication costs. Inference in such distributed setting would allow new applications, but requires balancing a triple trade-off between computation cost, communication bandwidth, and model accuracy. We explore that trade-off by studying the compressibility of representations at different stages of MobileNetV2, showing those results agree with theoretical intuitions about deep learning, and that an optimal splitting layer for network can be found with a simple PCA-based compression scheme.

State-of-the-art solutions in the areas of "Language Modelling & Generating Text", "Speech Recognition", "Generating Image Descriptions" or "Video Tagging" have been using Recurrent Neural Networks as the foundation for their approaches. Understanding the underlying concepts is therefore of tremendous importance if we want to keep up with recent or upcoming publications in those areas. In this work we give a short overview over some of the most important concepts in the realm of Recurrent Neural Networks which enables readers to easily understand the fundamentals such as but not limited to "Backpropagation through Time" or "Long Short-Term Memory Units" as well as some of the more recent advances like the "Attention Mechanism" or "Pointer Networks". We also give recommendations for further reading regarding more complex topics where it is necessary.

In this paper, we present a deep neural network (DNN) training approach called the "DeepMimic" training method. Enormous amounts of data are available nowadays for training usage. Yet, only a tiny portion of these data is manually labeled, whereas almost all of the data are unlabeled. The training approach presented utilizes, in a most simplified manner, the unlabeled data to the fullest, in order to achieve remarkable (classification) results. Our DeepMimic method uses a small portion of labeled data and a large amount of unlabeled data for the training process, as expected in a real-world scenario. It consists of a mentor model and a student model. Employing a mentor model trained on a small portion of the labeled data and then feeding it only with unlabeled data, we show how to obtain a (simplified) student model that reaches the same accuracy and loss as the mentor model, on the same test set, without using any of the original data labels in the training of the student model. Our experiments demonstrate that even on challenging classification tasks the student network architecture can be simplified significantly with a minor influence on the performance, i.e., we need not even know the original network architecture of the mentor. In addition, the time required for training the student model to reach the mentor's performance level is shorter, as a result of a simplified architecture and more available data. The proposed method highlights the disadvantages of regular supervised training and demonstrates the benefits of a less traditional training approach.

GROUND TRUTH SIMULA TION FOR DEEP LEARNING CLASSIFICA TION OF MID-RESOLUTION VENUS IMAGES VIA UNMIXING OF HIGH-RESOLUTION HYPERSPECTRAL FENIX DA T A Ido Faran, Nathan S. Netanyahu Eli (Omid) David Bar-Ilan University Dept. of Computer Science ramat-gan 5290002, Israel Maxim Shoshany, Fadi Kizel Jisung Geba Chang, Ronit Rud Technion Israel Institute of Technology Faculty of Civil and Environmental Engineering Haifa 3200003, Israel ABSTRACT Training a deep neural network for classification constitutes a major problem in remote sensing due to the lack of adequate field data. Acquiring high-resolution ground truth (GT) by human interpretation is both cost-ineffective and inconsistent. We propose, instead, to utilize high-resolution, hyperspectral images for solving this problem, by unmixing these images to obtain reliable GT for training a deep network. Specifically, we simulate GT from high-resolution, hyperspectral FENIX images, and use it for training a convolutional neural network (CNN) for pixel-based classification. We show how the model can be transferred successfully to classify new mid-resolution VENµ S imagery.

Intermittent demand, where demand occurrences appear sporadically in time, is a common and challenging problem in forecasting. In this paper, we first make the connections between renewal processes, and a collection of current models used for intermittent demand forecasting. We then develop a set of models that benefit from recurrent neural networks to parameterize conditional interdemand time and size distributions, building on the latest paradigm in "deep" temporal point processes. We present favorable empirical findings on discrete and continuous time intermittent demand data, validating the practical value of our approach.

Massive electronic health records (EHRs) enable the success of learning accurate patient representations to support various predictive health applications. In contrast, doctor representation was not well studied despite that doctors play pivotal roles in healthcare. How to construct the right doctor representations? How to use doctor representation to solve important health analytic problems? In this work, we study the problem on {\it clinical trial recruitment}, which is about identifying the right doctors to help conduct the trials based on the trial description and patient EHR data of those doctors. We propose doctor2vec which simultaneously learns 1) doctor representations from EHR data and 2) trial representations from the description and categorical information about the trials. In particular, doctor2vec utilizes a dynamic memory network where the doctor's experience with patients are stored in the memory bank and the network will dynamically assign weights based on the trial representation via an attention mechanism. Validated on large real-world trials and EHR data including 2,609 trials, 25K doctors and 430K patients, doctor2vec demonstrated improved performance over the best baseline by up to $8.7\%$ in PR-AUC. We also demonstrated that the doctor2vec embedding can be transferred to benefit data insufficiency settings including trial recruitment in less populated/newly explored country with $13.7\%$ improvement or for rare diseases with $8.1\%$ improvement in PR-AUC.

One promising research direction is to view deep neural networks through the lens of information theory (Tishby and Zaslavsky, 2015). Abstractly, deep connections exist between the information a learning algorithm extracts and its generalization capabilities (Bassily et al., 2017; Banerjee, 2006). Inspired by these general results, recent papers have attempted to measure information-theoretic quantities in ordinary deterministic neural networks (Shwartz-Ziv and Tishby, 2017; Achille and Soatto, 2017; Achille and Soatto, 2019). Both practical and theoretical problems arise in the deterministic case (Amjad and Geiger, 2018; Saxe et al., 2018; Kolchinsky et al., 2018). These difficulties stem from the fact that mutual information (MI) is reparameterization independent (Cover and Thomas, 2012). 1 One workaround is to make a network explicitly stochastic, either in its activations (Alemi et al., 2016) or its weights (Achille and Soatto, 2017).

LSTMs promise much to financial time-series analysis, temporal and cross-sectional inference, but we find they do not deliver in a real-world financial management task. We examine an alternative called Continual Learning (CL), a memory-augmented approach, which can provide transparent explanations; which memory did what and when. This work has implications for many financial applications including to credit, time-varying fairness in decision making and more. We make three important new observations. Firstly, as well as being more explainable, time-series CL approaches outperform LSTM and a simple sliding window learner (feed-forward neural net (FFNN)). Secondly, we show that CL based on a sliding window learner (FFNN) is more effective than CL based on a sequential learner (LSTM). Thirdly, we examine how real-world, time-series noise impacts several similarity approaches used in CL memory addressing. We provide these insights using an approach called Continual Learning Augmentation (CLA) tested on a complex real world problem; emerging market equities investment decision making. CLA provides a test-bed as it can be based on different types of time-series learner, allowing testing of LSTM and sliding window (FFNN) learners side by side. CLA is also used to test several distance approaches used in a memory recall-gate: euclidean distance (ED), dynamic time warping (DTW), auto-encoder (AE) and a novel hybrid approach, warp-AE. We find CLA out-performs simple LSTM and FFNN learners and CLA based on a sliding window (CLA-FFNN) out-performs a LSTM (CLA-LSTM) implementation. While for memory-addressing, ED under-performs DTW and AE but warp-AE shows the best overall performance in a real-world financial task.

The retail banking services are one of the pillars of the modern economic growth. However, the evolution of the client's habits in modern societies and the recent European regulations promoting more competition mean the retail banks will encounter serious challenges for the next few years, endangering their activities. They now face an impossible compromise: maximizing the satisfaction of their hyper-connected clients while avoiding any risk of default and being regulatory compliant. Therefore, advanced and novel research concepts are a serious game-changer to gain a competitive advantage. In this context, we investigate in this thesis different concepts bridging the gap between persistent homology, neural networks, recommender engines and reinforcement learning with the aim of improving the quality of the retail banking services. Our contribution is threefold. First, we highlight how to overcome insufficient financial data by generating artificial data using generative models and persistent homology. Then, we present how to perform accurate financial recommendations in multi-dimensions. Finally, we underline a reinforcement learning model-free approach to determine the optimal policy of money management based on the aggregated financial transactions of the clients. Our experimental data sets, extracted from well-known institutions where the privacy and the confidentiality of the clients were not put at risk, support our contributions. In this work, we provide the motivations of our retail banking research project, describe the theory employed to improve the financial services quality and evaluate quantitatively and qualitatively our methodologies for each of the proposed research scenarios.

Joint extraction of entities and relations has received significant attention due to its potential of providing higher performance for both tasks. Among existing methods, CopyRE is effective and novel, which uses a sequence-to-sequence framework and copy mechanism to directly generate the relation triplets. However, it suffers from two fatal problems. The model is extremely weak at differing the head and tail entity, resulting in inaccurate entity extraction. It also cannot predict multi-token entities (e.g. \textit{Steven Jobs}). To address these problems, we give a detailed analysis of the reasons behind the inaccurate entity extraction problem, and then propose a simple but extremely effective model structure to solve this problem. In addition, we propose a multi-task learning framework equipped with copy mechanism, called CopyMTL, to allow the model to predict multi-token entities. Experiments reveal the problems of CopyRE and show that our model achieves significant improvement over the current state-of-the-art method by 9% in NYT and 16% in WebNLG (F1 score). Our code is available at https://github.com/WindChimeRan/CopyMTL