Summarization of long sequences into a concise statement is a core problem in natural language processing, requiring non-trivial understanding of the input. Based on the promising results of graph neural networks on highly structured data, we develop a framework to extend existing sequence encoders with a graph component that can reason about long-distance relationships in weakly structured data such as text. In an extensive evaluation, we show that the resulting hybrid sequence-graph models outperform both pure sequence models as well as pure graph models on a range of summarization tasks.

Machine learning has been applied to a broad range of applications and some of them are available online as application programming interfaces (APIs) with either free (trial) or paid subscriptions. In this paper, we study adversarial machine learning in the form of back-box attacks on online classifier APIs. We start with a deep learning based exploratory (inference) attack, which aims to build a classifier that can provide similar classification results (labels) as the target classifier. To minimize the difference between the labels returned by the inferred classifier and the target classifier, we show that the deep learning based exploratory attack requires a large number of labeled training data samples. These labels can be collected by calling the online API, but usually there is some strict rate limitation on the number of allowed API calls. To mitigate the impact of limited training data, we develop an active learning approach that first builds a classifier based on a small number of API calls and uses this classifier to select samples to further collect their labels. Then, a new classifier is built using more training data samples. This updating process can be repeated multiple times. We show that this active learning approach can build an adversarial classifier with a small statistical difference from the target classifier using only a limited number of training data samples. We further consider evasion and causative (poisoning) attacks based on the inferred classifier that is built by the exploratory attack. Evasion attack determines samples that the target classifier is likely to misclassify, whereas causative attack provides erroneous training data samples to reduce the reliability of the re-trained classifier. The success of these attacks show that adversarial machine learning emerges as a feasible threat in the realistic case with limited training data.

Deep neural networks use stacked layers of feature detectors to repeatedly transform the input data, so that structurally different classes of input become well separated in the final layer. While the method has turned out extremely powerful in many applications, its success depends critically on the correct choice of hyperparameters, in particular the number of network layers. Here, we introduce a new measure, called the generalized discrimination value (GDV), which quantifies how well different object classes separate in each layer. Due to its definition, the GDV is invariant to translation and scaling of the input data, independent of the number of features, as well as independent of the number and permutation of the neurons within a layer. We compute the GDV in each layer of a Deep Belief Network that was trained unsupervised on the MNIST data set. Strikingly, we find that the GDV first improves with each successive network layer, but then gets worse again beyond layer 30, thus indicating the optimal network depth for this data classification task. Our further investigations suggest that the GDV can serve as a universal tool to determine the optimal number of layers in deep neural networks for any type of input data.

Despite numerous state-of-the-art applications of Deep Neural Networks (DNNs) in a wide range of real-world tasks, two major challenges hinder further advances in DNNs: hyperparameter optimization and lack of computing power. DNNs become increasingly difficult to train and deploy as they grow in size due to both computational intensity and the large resulting memory footprint. Recent efforts show that quantizing the weights and activations of DNN layers to lower bitwidths takes a significant step toward reducing memory bandwidth and power consumption by using limited computing resources. This paper builds upon the algorithmic insight that the bitwidth of operations in DNNs can be reduced without compromising their classification accuracy. While the use of eight-bit weights and activations during inference maintains the accuracy in most cases, lower bitwidths can achieve the same accuracy while utilizing less power. However, deep quantization (quantizing bitwidths below eight) while maintaining accuracy requires a great deal of trial-and-error, fine-tuning as well as retraining. By formulating quantization bitwidth as a hyperparameter in the optimization problem of selecting the bitwidth, we tackle this issue by leveraging a state-of-the-art policy gradient based Reinforcement Learning (RL) algorithm called Proximal Policy Optimization [10] (PPO), to efficiently explore a large design space of DNN quantization. The proposed technique also opens up the possibility of performing heterogeneous quantization of the network (e.g., quantizing each layer to different bitwidth) as the RL agent learns the sensitivity of each layer with respect to accuracy in order to perform quantization of the entire network. We evaluated our method on several neural networks including MNIST, CIFAR10, SVHN and the RL agent quantizes these networks to average bitwidths of 2.25, 5 and 4 respectively with less than 0.3% accuracy loss in all cases.

Abstract--Recently, word embedding algorithms have been applied to map the entities of recommender systems, such as users and items, to new feature spaces using textual elementcontext relations among them. Unlike many other domains, this approach has not achieved a desired performance in collaborative filtering problems, probably due to unavailability of appropriate textual data. In this paper we propose a new recommendation framework, called GEMRank that can be applied when the useritem matrix is the sole available souce of information. It uses the concept of profile co-occurrence for defining relations among entities and applies a factorization method for embedding the users and items. GEMRank then feeds the extracted representations to a neural network model to predict user-item like/dislike relations which the final recommendations are made based on. We evaluated GEMRank in an extensive set of experiments against state of the art recommendation methods. The results show that GEMRank significantly outperforms the baseline algorithms in a variety of data sets with different degrees of density. Recommendation Systems help users to find relevant items based on their preferences. Many prominent recommendation systems are using Collaborative Filtering (CF) for making recommendations ( [1]).

We present a novel approach to leverage large unlabeled datasets by pre-training state-of-the-art deep neural networks on randomly-labeled datasets. Specifically, we train the neural networks to memorize arbitrary labels for all the samples in a dataset and use these pre-trained networks as a starting point for regular supervised learning. Our assumption is that the "memorization infrastructure" learned by the network during the random-label training proves to be beneficial for the conventional supervised learning as well. We test the effectiveness of our pre-training on several video action recognition datasets (HMDB51, UCF101, Kinetics) by comparing the results of the same network with and without the random label pre-training. Our approach yields an improvement - ranging from 1.5% on UCF-101 to 5% on Kinetics - in classification accuracy, which calls for further research in this direction.

Epilepsy is the most common neurological disorder and an accurate forecast of seizures would help to overcome the patient's uncertainty and helplessness. In this contribution, we present and discuss a novel methodology for the classification of intracranial electroencephalography (iEEG) for seizure prediction. Contrary to previous approaches, we categorically refrain from an extraction of hand-crafted features and use a convolutional neural network (CNN) topology instead for both the determination of suitable signal characteristics and the binary classification of preictal and interictal segments. Three different models have been evaluated on public datasets with long-term recordings from four dogs and three patients. Overall, our findings demonstrate the general applicability. In this work we discuss the strengths and limitations of our methodology.

Recent works have shown that it is possible to train models that are verifiably robust to norm-bounded adversarial perturbations. While these recent methods show promise, they remain hard to scale and difficult to tune. This paper investigates how interval bound propagation (IBP) using simple interval arithmetic can be exploited to train verifiably robust neural networks that are surprisingly effective. While IBP itself has been studied in prior work, our contribution is in showing that, with an appropriate loss and careful tuning of hyper-parameters, verified training with IBP leads to a fast and stable learning algorithm. We compare our approach with recent techniques, and train classifiers that improve on the state-of-the-art in single-model adversarial robustness: we reduce the verified error rate from 3.67% to 2.23% on MNIST (with $\ell_\infty$ perturbations of $\epsilon = 0.1$), from 19.32% to 8.05% on MNIST (at $\epsilon = 0.3$), and from 78.22% to 72.91% on CIFAR-10 (at $\epsilon = 8/255$).

Separating an audio scene into isolated sources is a fundamental problem in computer audition, analogous to image segmentation in visual scene analysis. Source separation systems based on deep learning are currently the most successful approaches for solving the underdetermined separation problem, where there are more sources than channels. Traditionally, such systems are trained on sound mixtures where the ground truth decomposition is already known. Since most real-world recordings do not have such a decomposition available, this limits the range of mixtures one can train on, and the range of mixtures the learned models may successfully separate. In this work, we use a simple blind spatial source separation algorithm to generate estimated decompositions of stereo mixtures. These estimates, together with a weighting scheme in the time-frequency domain, based on confidence in the separation quality, are used to train a deep learning model that can be used for single-channel separation, where no source direction information is available. This demonstrates how a simple cue such as the direction of origin of source can be used to bootstrap a model for source separation that can be used in situations where that cue is not available.

Connectionist temporal classification (CTC) training criterion provides an alternative acoustic model (AM) training strategy for automatic speech recognition in an end-to-end fashion. Although CTC criterion benefits acoustic modeling without needs of time-aligned phonetics transcription, it remains in need of efforts of tweaking to convergence, especially in the resource-constrained scenario. In this paper, we proposed to improve CTC training by incorporating acoustic landmarks. We tailored a new set of acoustic landmarks to help CTC training converge more quickly while also reducing recognition error rates. We leveraged new target label sequences mixed with both phone and manner changes to guide CTC training. Experiments on TIMIT demonstrated that CTC based acoustic models converge faster and smoother significantly when they are augmented by acoustic landmarks. The models pretrained with mixed target labels can be finetuned furthermore, which reduced phone error rate by 8.72% on TIMIT. The consistent performance gain is also observed on reduced TIMIT and WSJ as well, in which case, we are the first to succeed in testing the effectiveness of acoustic landmark theory on mid-sized ASR tasks.