As you read this essay, you understand each word based on your understanding of previous words. You don't throw everything away and start thinking from scratch again. Traditional neural networks can't do this, and it seems like a major shortcoming. For example, imagine you want to classify what kind of event is happening at every point in a movie. It's unclear how a traditional neural network could use its reasoning about previous events in the film to inform later ones. Recurrent neural networks address this issue.

The ability to stratify patients by cardiovascular risk is essential for identifying those likely to suffer a heart attack, stroke, or other heart disease in the future. High-risk patients can then take steps to improve their cardiovascular health. Doctors typically take into account a variety of risk factors: demographics such as age, sex and ethnicity; daily behaviors like exercise, smoking status and diet; as well as results from blood pressure and cholesterol tests. As a simple alternative to the traditional patient questionnaire and blood tests, a team of researchers from Google Research and the Stanford School of Medicine have developed deep learning models to predict cardiovascular risk factors from photographs of the back of the retina. Since these retinal fundus images are already collected for diabetic eye disease screening, this initial study suggests that deep learning could uncover additional information that could be further leveraged for preventative health.

In this work, we focus on the task of generating natural language descriptions from a structured table of facts containing fields (such as nationality, occupation, etc) and values (such as Indian, actor, director, etc). One simple choice is to treat the table as a sequence of fields and values and then use a standard seq2seq model for this task. However, such a model is too generic and does not exploit task-specific characteristics. For example, while generating descriptions from a table, a human would attend to information at two levels: (i) the fields (macro level) and (ii) the values within the field (micro level). Further, a human would continue attending to a field for a few timesteps till all the information from that field has been rendered and then never return back to this field (because there is nothing left to say about it). To capture this behavior we use (i) a fused bifocal attention mechanism which exploits and combines this micro and macro level information and (ii) a gated orthogonalization mechanism which tries to ensure that a field is remembered for a few time steps and then forgotten. We experiment with a recently released dataset which contains fact tables about people and their corresponding one line biographical descriptions in English. In addition, we also introduce two similar datasets for French and German. Our experiments show that the proposed model gives 21% relative improvement over a recently proposed state of the art method and 10% relative improvement over basic seq2seq models. The code and the datasets developed as a part of this work are publicly available.

Modern deep neural network training is typically based on mini-batch stochastic gradient optimization. While the use of large mini-batches increases the available computational parallelism, small batch training has been shown to provide improved generalization performance and allows a significantly smaller memory footprint, which might also be exploited to improve machine throughput. In this paper, we review common assumptions on learning rate scaling and training duration, as a basis for an experimental comparison of test performance for different mini-batch sizes. We adopt a learning rate that corresponds to a constant average weight update per gradient calculation (i.e., per unit cost of computation), and point out that this results in a variance of the weight updates that increases linearly with the mini-batch size $m$. The collected experimental results for the CIFAR-10, CIFAR-100 and ImageNet datasets show that increasing the mini-batch size progressively reduces the range of learning rates that provide stable convergence and acceptable test performance. On the other hand, small mini-batch sizes provide more up-to-date gradient calculations, which yields more stable and reliable training. The best performance has been consistently obtained for mini-batch sizes between $m = 2$ and $m = 32$, which contrasts with recent work advocating the use of mini-batch sizes in the thousands.

Variational methods for revealing visual concepts learned by convolutional neural networks have gained significant attention during the last years. Being based on noisy gradients obtained via back-propagation such methods require the application of regularization strategies. We present a mathematical framework unifying previously employed regularization methods. Within this framework, we propose a novel technique based on Sobolev gradients which can be implemented via convolutions and does not require specialized numerical treatment, such as total variation regularization. The experiments performed on feature inversion and activation maximization demonstrate the benefit of a unified approach to regularization, such as sharper reconstructions via the proposed Sobolev filters and a better control over reconstructed scales.

While deep neural networks have proven to be a powerful tool for many recognition and classification tasks, their stability properties are still not well understood. In the past, image classifiers have been shown to be vulnerable to so-called adversarial attacks, which are created by additively perturbing the correctly classified image. In this paper, we propose the ADef algorithm to construct a different kind of adversarial attack created by iteratively applying small deformations to the image, found through a gradient descent step. We demonstrate our results on MNIST with a convolutional neural network and on ImageNet with Inception-v3 and ResNet-101.

In this paper, we apply neural networks into digital marketing world for the purpose of better targeting the potential customers. To do so, we model the customer online behaviours using dedicated neural network architectures. Starting from user searched keywords in a search engine to the landing page and different following pages, until the user left the site, we model the whole visited journey with a Recurrent Neural Network (RNN), together with Convolution Neural Networks (CNN) that can take into account of the semantic meaning of user searched keywords and different visited page names. With such model, we use Monte Carlo simulation to estimate the conversion rates of each potential customer in the future visiting. We believe our concept and the preliminary promising results in this paper enable the use of largely available customer online behaviours data for advanced digital marketing analysis.

Deep neural networks have proved to be effective function approximators in Reinforcement Learning (RL). Significant progress is seen in many RL problems ranging from board games like Go (Silver et al., 2016, 2017b), Chess and Shogi (Silver et al., 2017a), video games like Atari (Mnih et al., 2015) and StarCraft (Vinyals et al., 2017), to real world robotics and control tasks (Lillicrap et al., 2016). Most of these successes are due to improved training algorithms, carefully designed neural network architectures and powerful hardware. For example, in AlphaZero (Silver et al., 2017a), 5,000 1st-generation TPUs and 64 2nd-generation TPUs are used during self-play based training of agents with deep residual networks (He et al., 2016). On the other hand, learning with high-capacity models and long stretched training time on powerful devices could lead to potential risk of overfitting (Hardt et al., 2016; Lin et al., 2016). As a fundamental tradeoff in machine learning, preventing overfitting by properly controlling or regularizing the training is key to out-of-sample generalization. Studies of overfitting could be performed from the theory side, where generalization guarantees are derived for specific learning algorithms; or from the practice side, where carefully designed experimental protocols like cross validation are used as proxy to certify the generalization performance. Unfortunately, in the regime of deep RL, systematic studies of generalization behaviors from either theoretical or empirical perspectives are falling behind the rapid progresses from the algorithm development and application side. The current situation not only makes it difficult to understand the test behaviors like the vulnerabilities to potential adversarial attacks (Huang et al., 2017), but also renders some results difficult to reproduce or compare (Henderson et al., 2017; Machado et al., 2017).

Machine learning is used to compute achievable information rates (AIRs) for a simplified fiber channel. The approach jointly optimizes the input distribution (constellation shaping) and the auxiliary channel distribution to compute AIRs without explicit channel knowledge in an end-to-end fashion. Proposed Autoencoder Structure In machine learning, an AE is a neural network (NN) which consists of two parts: an encoder maps an input s (e.g., an image) to a lowerdimensional representation or code and a decoder attempts to reconstruct the input from the code. The AE structure used in this paper is shown in Figure 1 and will be described in the following. The goal is to transmit a message s chosen from a set of M possible messages {1, 2,..., M} M. Following 9 In this paper, we develop an AE for a simplified memoryless fiber channel model.

In this paper, we propose a novel unsupervised learning method to learn the brain dynamics using a deep learning architecture named residual D-net. As it is often the case in medical research, in contrast to typical deep learning tasks, the size of the resting-state functional Magnetic Resonance Image (rs-fMRI) datasets for training is limited. Thus, the available data should be very efficiently used to learn the complex patterns underneath the brain connectivity dynamics. To address this issue, we use residual connections to alleviate the training complexity through recurrent multi-scale representation. We conduct two classification tasks to differentiate early and late stage Mild Cognitive Impairment (MCI) from Normal healthy Control (NC) subjects. The experiments verify that our proposed residual D-net indeed learns the brain connectivity dynamics, leading to significantly higher classification accuracy compared to previously published techniques.