In this paper, we explore the ways to improve POS-tagging using various types of auxiliary losses and different word representations. As a baseline, we utilized a BiLSTM tagger, which is able to achieve state-of-the-art results on the sequence labelling tasks. We developed a new method for character-level word representation using feedforward neural network. Such representation gave us better results in terms of speed and performance of the model. We also applied a novel technique of pretraining such word representations with existing word vectors. Finally, we designed a new variant of auxiliary loss for sequence labelling tasks: an additional prediction of the neighbour labels. Such loss forces a model to learn the dependencies in-side a sequence of labels and accelerates the process of training. We test these methods on English and Russian languages.

The stock market is a nonlinear, nonstationary, dynamic, and complex system. There are several factors that affect the stock market conditions, such as news, social media, expert opinion, political transitions, and natural disasters. In addition, the market must also be able to handle the situation of illegal insider trading, which impacts the integrity and value of stocks. Illegal insider trading occurs when trading is performed based on non-public (private, leaked, tipped) information (e.g., new product launch, quarterly financial report, acquisition or merger plan) before the information is made public. Preventing illegal insider trading is a priority of the regulatory authorities (e.g., SEC) as it involves billions of dollars, and is very difficult to detect. In this work, we present different types of insider trading approaches, techniques and our proposed approach for detecting and predicting insider trader using a deep-learning based approach combined with discrete signal processing on time series data.

Manually labeled corpora are expensive to create and often not available for low-resource languages or domains. Automatic labeling approaches are an alternative way to obtain labeled data in a quicker and cheaper way. However, these labels often contain more errors which can deteriorate a classifier's performance when trained on this data. We propose a noise layer that is added to a neural network architecture. This allows modeling the noise and train on a combination of clean and noisy data. We show that in a low-resource NER task we can improve performance by up to 35% by using additional, noisy data and handling the noise.

Deep Neural Networks (DNNs) have become increasingly popular in computer vision, natural language processing, and other areas. However, training and fine-tuning a deep learning model is computationally intensive and time-consuming. We propose a new method to improve the performance of nearly every model including pre-trained models. The proposed method uses an ensemble approach where the networks in the ensemble are constructed by reassigning model parameter values based on the probabilistic distribution of these parameters, calculated towards the end of the training process. For pre-trained models, this approach results in an additional training step (usually less than one epoch). We perform a variety of analysis using the MNIST dataset and validate the approach with a number of DNN models using pre-trained models on the ImageNet dataset.

Fairness is a critical trait in decision making. As machine-learning models are increasingly being used in sensitive application domains (e.g. education and employment) for decision making, it is crucial that the decisions computed by such models are free of unintended bias. But how can we automatically validate the fairness of arbitrary machine-learning models? For a given machine-learning model and a set of sensitive input parameters, our AEQUITAS approach automatically discovers discriminatory inputs that highlight fairness violation. At the core of AEQUITAS are three novel strategies to employ probabilistic search over the input space with the objective of uncovering fairness violation. Our AEQUITAS approach leverages inherent robustness property in common machine-learning models to design and implement scalable test generation methodologies. An appealing feature of our generated test inputs is that they can be systematically added to the training set of the underlying model and improve its fairness. To this end, we design a fully automated module that guarantees to improve the fairness of the underlying model. We implemented AEQUITAS and we have evaluated it on six state-of-the-art classifiers, including a classifier that was designed with fairness constraints. We show that AEQUITAS effectively generates inputs to uncover fairness violation in all the subject classifiers and systematically improves the fairness of the respective models using the generated test inputs. In our evaluation, AEQUITAS generates up to 70% discriminatory inputs (w.r.t. the total number of inputs generated) and leverages these inputs to improve the fairness up to 94%.

Noname manuscript No. (will be inserted by the editor) Abstract Routing questions in Community Question Answer services (CQAs) such as Stack Exchange sites is a well-studied problem. Yet, cold-start - a phenomena observed when a new question is posted is not well addressed by existing approaches. Additionally, cold questions posted by new askers present significant challenges to state-of-the-art approaches. We propose ColdRoute to address these challenges. ColdRoute is able to handle the task of routing cold questions posted by new or existing askers to matching experts. Specifically, we use Factorization Machines on the one-hot encoding of critical features such as question tags and compare our approach to well-studied techniques such as CQARank and semantic matching (LDA, BoW, and Doc2Vec). Using data from eight stack exchange sites, we are able to improve upon the routing metrics (Precision@1, Accuracy, MRR) over the state-of-the-art models such as semantic matching by 159.5%,31.84%, Keywords question routing · expert finding · cold-start problem · question answering services 1 Introduction Nowadays, the Community-based question answering sites (CQAs) such as Stack Overflow, Stack Exchange Sites, and Quora, which enable people to post questions and answers in various domains [Yang et al., 2013] have accumulated millions Aniket Chakrabarti Microsoft (work done while at The Ohio State University) Email: chakrabarti.14@osu.edu 2 Jiankai Sun et al. One important task in CQAs is to make recommendations for new questions (routing questions), that fall in three scenarios: 1) find experts. In this paper, we focus on the problem of expert finding [Xu et al., 2012,Zhao et al., 2013,Yang et al., 2013, Fang et al., 2016,Zhao et al., 2016,Zhao et al., 2017], which is to choose the right experts for answering questions posted by users in Stack Exchange, which is a network of question-and-answer (Q&A) websites containing topics in various fields. Each Stack Exchange site covers a specific topic. Usually there are two types of questions in CQAs - resolved (questions with answers) and newly posted questions (questions that have not received any answers).

We release two artificial datasets, Simulated Flying Shapes and Simulated Planar Manipulator that allow to test the learning ability of video processing systems. In particular, the dataset is meant as a tool which allows to easily assess the sanity of deep neural network models that aim to encode, reconstruct or predict video frame sequences. The datasets each consist of 90000 videos. The Simulated Flying Shapes dataset comprises scenes showing two objects of equal shape (rectangle, triangle and circle) and size in which one object approaches its counterpart. The Simulated Planar Manipulator shows a 3-DOF planar manipulator that executes a pick-and-place task in which it has to place a size-varying circle on a squared platform. Different from other widely used datasets such as moving MNIST [1], [2], the two presented datasets involve goal-oriented tasks (e.g. the manipulator grasping an object and placing it on a platform), rather than showing random movements. This makes our datasets more suitable for testing prediction capabilities and the learning of sophisticated motions by a machine learning model. This technical document aims at providing an introduction into the usage of both datasets.

Deep neural networks (DNN) are increasingly being accelerated on application-specific hardware such as the Google TPU designed especially for deep learning. Timing speculation is a promising approach to further increase the energy efficiency of DNN accelerators. Architectural exploration for timing speculation requires detailed gate-level timing simulations that can be time-consuming for large DNNs that execute millions of multiply-and-accumulate (MAC) operations. In this paper we propose FATE, a new methodology for fast and accurate timing simulations of DNN accelerators like the Google TPU. FATE proposes two novel ideas: (i) DelayNet, a DNN based timing model for MAC units; and (ii) a statistical sampling methodology that reduces the number of MAC operations for which timing simulations are performed. We show that FATE results in between 8 times-58 times speed-up in timing simulations, while introducing less than 2% error in classification accuracy estimates. We demonstrate the use of FATE by comparing to conventional DNN accelerator that uses 2's complement (2C) arithmetic with an alternative implementation that uses signed magnitude representations (SMR). We show that that the SMR implementation provides 18% more energy savings for the same classification accuracy than 2C, a result that might be of independent interest.

New deep learning techniques have enabled us to analyze video footage more quickly and cheaply than ever before, and a larger number of companies in Japan, America, and China are developing products with resembling capabilities. Similar features are making their way into home security cameras, with companies like Amazon and Nest offering rudimentary AI analysis. This sort of automated surveillance is only going to become more common in the future, with researchers working on advanced analysis like spotting violent behavior in crowds, and tech companies selling tools like facial recognition to law enforcement. A new AI security cam is designed to help shop owners in Japan spot potential shoplifters. It uses open source technology developed by Carnegie Mellon University to scan live video streams and estimate the poses of any bodies it can see. The security camera, called the "AI Guardman", was built by Japanese telecom giant NTT East and startup Earth Eyes Corp.

Given how artificial intelligence is a buzzing topic, it has sparked a slew of beginner-friendly introductory resources that clear the general concepts from this very broad topic. And for most newcomers, the most interesting topic in AI is Deep Learning. In fact, Google's Python-based Deep Learning framework Tensorflow has helped many a developer get up to speed with the technical concepts. Besides videos and free online courses, you must also have a reading list that helps you cover the math and statistics behind the algorithms. While YouTube videos remain the main learning source and a key starting point for beginners, there is a slew of resources, especially books that can help cement fundamental concepts.