As mobile devices become more and more popular, mobile gaming has emerged as a promising market with billion-dollar revenues. A variety of mobile game platforms and services have been developed around the world. A critical challenge for these platforms and services is to understand the churn behavior in mobile games, which usually involves churn at micro level (between an app and a specific user) and macro level (between an app and all its users). Accurate micro-level churn prediction and macro-level churn ranking will benefit many stakeholders such as game developers, advertisers, and platform operators. In this paper, we present the first large-scale churn analysis for mobile games that supports both micro-level churn prediction and macro-level churn ranking. For micro-level churn prediction, in view of the common limitations of the state-of-the-art methods built upon traditional machine learning models, we devise a novel semi-supervised and inductive embedding model that jointly learns the prediction function and the embedding function for user-app relationships. We model these two functions by deep neural networks with a unique edge embedding technique that is able to capture both contextual information and relationship dynamics. We also design a novel attributed random walk technique that takes into consideration both topological adjacency and attribute similarities. To address macro-level churn ranking, we propose to construct a relationship graph with estimated micro-level churn probabilities as edge weights and adapt link analysis algorithms on the graph. We devise a simple algorithm SimSum and adapt two more advanced algorithms PageRank and HITS. The performance of our solutions for the two-level churn analysis problems is evaluated on real-world data collected from the Samsung Game Launcher platform.

Machine-learning models have demonstrated great success in learning complex patterns that enable them to make predictions about unobserved data. In addition to using models for prediction, the ability to interpret what a model has learned is receiving an increasing amount of attention. However, this increased focus has led to considerable confusion about the notion of interpretability. In particular, it is unclear how the wide array of proposed interpretation methods are related, and what common concepts can be used to evaluate them. We aim to address these concerns by defining interpretability in the context of machine learning and introducing the Predictive, Descriptive, Relevant (PDR) framework for discussing interpretations. The PDR framework provides three overarching desiderata for evaluation: predictive accuracy, descriptive accuracy and relevancy, with relevancy judged relative to a human audience. Moreover, to help manage the deluge of interpretation methods, we introduce a categorization of existing techniques into model-based and post-hoc categories, with sub-groups including sparsity, modularity and simulatability. To demonstrate how practitioners can use the PDR framework to evaluate and understand interpretations, we provide numerous real-world examples. These examples highlight the often under-appreciated role played by human audiences in discussions of interpretability. Finally, based on our framework, we discuss limitations of existing methods and directions for future work. We hope that this work will provide a common vocabulary that will make it easier for both practitioners and researchers to discuss and choose from the full range of interpretation methods.

Abstract--Previous attempts at music artist classification use frame-level audio features which summarize frequency content within short intervals of time. Comparatively, more recent music information retrieval tasks take advantage of temporal structure in audio spectrograms using deep convolutional and recurrent models. This paper revisits artist classification with this new framework and empirically explores the impacts of incorporating temporal structure in the feature representation. To this end, an established classification architecture, a Convolutional Recurrent Neural Network (CRNN), is applied to the artist20 music artist identification dataset under a comprehensive set of conditions. These include audio clip length, which is a novel contribution in this work, and previously identified considerations such as dataset split and feature-level. Our results improve upon baseline works, verify the influence of the production details on classification performance and demonstrate the tradeoffs between sample length and training set size. The best performing model achieves an average F1-score of 0.937 across three independent trials which is a substantial improvement over the corresponding baseline under similar conditions. Finally, to showcase the effectiveness ofthe CRNN's feature extraction capabilities, we visualize audio samples at its bottleneck layer demonstrating that learned representations segment into clusters belonging to their respective artists. I. INTRODUCTION Music information retrieval (MIR) encompasses most audio analysis tasks such as genre classification, song identification, chord recognition, sound event detection, mood detection and feature extraction [1], [2].

Deep learning is very effective at jointly learning feature representations and classification models, especially when dealing with high dimensional input patterns. Probabilistic logic reasoning, on the other hand, is capable to take consistent and robust decisions in complex environments. The integration of deep learning and logic reasoning is still an open-research problem and it is considered to be the key for the development of real intelligent agents. This paper presents Deep Logic Models, which are deep graphical models integrating deep learning and logic reasoning both for learning and inference. Deep Logic Models create an end-to-end differentiable architecture, where deep learners are embedded into a network implementing a continuous relaxation of the logic knowledge. The learning process allows to jointly learn the weights of the deep learners and the meta-parameters controlling the high-level reasoning. The experimental results show that the proposed methodology overtakes the limitations of the other approaches that have been proposed to bridge deep learning and reasoning.

We introduce synthetic oversampling in anomaly detection for multi-feature sequence datasets based on autoencoders and generative adversarial networks. The first approach considers the use of an autoencoder in conjunction with standard oversampling methods to generate synthetic data that captures the sequential nature of the data. A different model uses generative adversarial networks to generate structure preserving synthetic data for the minority class. We also use generative adversarial networks on the majority class as an outlier detection method for novelty detection. We show that the use of generative adversarial network based synthetic data improves classification model performance on a variety of sequence data sets.

In this paper we propose a novel augmentation technique that improves not only the performance of deep neural networks on clean test data, but also significantly increases their robustness to random transformations, both affine and projective. Inspired by ManiFool, the augmentation is performed by a line-search manifold-exploration method that learns affine geometric transformations that lead to the misclassification on an image, while ensuring that it remains on the same manifold as the training data. This augmentation method populates any training dataset with images that lie on the border of the manifolds between two-classes and maximizes the variance the network is exposed to during training. Our method was thoroughly evaluated on the challenging tasks of fine-grained skin lesion classification from limited data, and breast tumor classification of mammograms. Compared with traditional augmentation methods, and with images synthesized by Generative Adversarial Networks our method not only achieves state-of-the-art performance but also significantly improves the network's robustness.

This paper proposes a method to reduce the training set of a deep learning (DL) model for testing purposes. The reduced training set allows faster model training than the original, full-sized training set and mimics the training characteristics of full-sized model training. The increasing number of software systems based on machine learning systems and learning components requires adequate testing techniques to ensure their functionality. Continuous Integration (CI) is a software development technique foriterative software development. Within CI cycles, changes to the software are integrated into the main software repository, built and tested. One major aspect of modern CI systems is to focus on short cycles times, that provide feedback about the change quality early after submitting the changes to version control. As deep learning (DL), as well as other machine learning systems are more prevalent and central to the software success, also their development and quality assurance is moved within the CI paradigm.

In this work, we propose a goal-driven collaborative task that contains language, vision, and action in a virtual environment as its core components. Specifically, we develop a Collaborative image-Drawing game between two agents, called CoDraw. Our game is grounded in a virtual world that contains movable clip art objects. The game involves two players: a Teller and a Drawer. The Teller sees an abstract scene containing multiple clip art pieces in a semantically meaningful configuration, while the Drawer tries to reconstruct the scene on an empty canvas using available clip art pieces. The two players communicate via two-way communication using natural language. We collect the CoDraw dataset of ~10K dialogs consisting of ~138K messages exchanged between human agents. We define protocols and metrics to evaluate the effectiveness of learned agents on this testbed, highlighting the need for a novel crosstalk condition which pairs agents trained independently on disjoint subsets of the training data for evaluation. We present models for our task, including simple but effective nearest-neighbor techniques and neural network approaches trained using a combination of imitation learning and goal-driven training. All models are benchmarked using both fully automated evaluation and by playing the game with live human agents.

We live in a world where, for better and for worse, we are constantly surrounded by deep learning algorithms. From social network filtering to driverless cars to movie recommendations, and from financial fraud detection to drug discovery to medical image processing (…is that bump cancer?), the field of deep learning influences our lives and our decisions every single day. In fact, you're probably reading this article right now because a deep learning algorithm thinks you should see it. If you're looking for the basics of deep learning, artificial neural networks, convolutional neural networks, (neural networks in general…), backpropagation, gradient descent, and more, you've come to the right place. In this series of articles, I'm going to explain what these concepts are as simply and comprehensibly as I can. If you get into this, there's an incredible amount of really in-depth information out there! I'll make sure to provide additional resources along the way for anyone who wants to swim a little deeper into these waters.

Long Short Term Memory networks, usually called "LSTMs", were introduced by Hochreiter and Schmiduber. These have widely been used for speech recognition, language modeling, sentiment analysis and text prediction. Before going deep into LSTM, we should first understand the need of LSTM which can be explained by the drawback of practical use of Recurrent Neural Network (RNN). So, lets start with RNN. Being human, when we watch a movie, we don't think from scratch every time while understanding any event.