Abstract-- We examine Deep Canonically Correlated LSTMs as a way to learn nonlinear transformations of variable length sequences and embed them into a correlated, fixed dimensional space. We use LSTMs to transform multi-view time-series data non-linearly while learning temporal relationships within the data. We then perform correlation analysis on the outputs of these neural networks to find a correlated subspace through which we get our final representation via projection. This work follows from previous work done on Deep Canonical Correlation (DCCA), in which deep feed-forward neural networks were used to learn nonlinear transformations of data while maximizing correlation. I. INTRODUCTION It is common in modern data sets to have multiple views of data collected of a phenomenon, for instance, a set of images and their captions in text, or audio and video data of the same event. If there exist labels, the views are conditionally uncorrelated on them, and it is typically assumed that noise sources between views are uncorrelated so that the representations are discriminating of the underlying semantic content. To distinguish it from multi-modal learning, multi-view learning trains a model or classifier for each view, the application of which depends on what data is available at test time. Typically it is desirable to find representations for each view that are predictive of - and predicted by - the other views so that if one view is not available at test time, it can serve to denoise the other views, or serve as a soft supervisor providing pseudo-labels. The benefits of training on multiple views include reduced sample complexity for prediction scenarios [1], relaxed separation conditions for clustering [2], among others. CCA techniques are used successfully across a wide array of downstream tasks (often unsupervised) from fMRI analysis [3], to retrieval, categorization, and clustering of text documents [4], [5], to acoustic feature learning [6]-[8].

Image denoising is always a challenging task in the field of computer vision and image processing. In this paper we have proposed an encoder-decoder model with direct attention, which is capable of denoising and reconstruct highly corrupted images. Our model is consisted of an encoder and a decoder, where encoder is a convolutional neural network and decoder is a multilayer Long Short-Term memory network. In the proposed model, the encoder reads an image and catches the abstraction of that image in a vector, where decoder takes that vector as well as the corrupted image to reconstruct a clean image. We have trained our model on MNIST handwritten digit database after making lower half of every image as black as well as adding noise top of that. After a massive destruction of the images where it is hard for a human to understand the content of those images, our model can retrieve that image with minimal error. Our proposed model has been compared with convolutional encoder-decoder, where our model has performed better at generating missing part of the images than convolutional auto encoder.

This paper first answers the question "why do the two most powerful techniques Dropout and Batch Normalization (BN) often lead to a worse performance when they are combined together?" in both theoretical and statistical aspects. Theoretically, we find that Dropout would shift the variance of a specific neural unit when we transfer the state of that network from train to test. However, BN would maintain its statistical variance, which is accumulated from the entire learning procedure, in the test phase. The inconsistency of that variance (we name this scheme as "variance shift") causes the unstable numerical behavior in inference that leads to more erroneous predictions finally, when applying Dropout before BN. Thorough experiments on DenseNet, ResNet, ResNeXt and Wide ResNet confirm our findings. According to the uncovered mechanism, we next explore several strategies that modifies Dropout and try to overcome the limitations of their combination by avoiding the variance shift risks.

Neural programming involves training neural networks to learn programs from data. Previous works have failed to achieve good generalization performance, especially on programs with high complexity or on large domains. This is because they mostly rely either on black-box function evaluations that do not capture the structure of the program, or on detailed execution traces that are expensive to obtain, and hence the training data has poor coverage of the domain under consideration. We present a novel framework that utilizes black-box function evaluations, in conjunction with symbolic expressions that integrate relationships between the given functions. We employ tree LSTMs to incorporate the structure of the symbolic expression trees. We use tree encoding for numbers present in function evaluation data, based on their decimal representation. We present an evaluation benchmark for this task to demonstrate our proposed model combines symbolic reasoning and function evaluation in a fruitful manner, obtaining high accuracies in our experiments. Our framework generalizes significantly better to expressions of higher depth and is able to fill partial equations with valid completions.

A widely studied non-deterministic polynomial time (NP) hard problem lies in finding a route between the two nodes of a graph. Often meta-heuristics algorithms such as $A^{*}$ are employed on graphs with a large number of nodes. Here, we propose a deep recurrent neural network architecture based on the Sequence-2-Sequence (Seq2Seq) model, widely used, for instance in text translation. Particularly, we illustrate that utilising a context vector that has been learned from two different recurrent networks enables increased accuracies in learning the shortest route of a graph. Additionally, we show that one can boost the performance of the Seq2Seq network by smoothing the loss function using a homotopy continuation of the decoder's loss function.

We propose a probabilistic model for interpreting gene expression levels that are observed through single-cell RNA sequencing. In the model, each cell has a low-dimensional latent representation. Additional latent variables account for technical effects that may erroneously set some observations of gene expression levels to zero. Conditional distributions are specified by neural networks, giving the proposed model enough flexibility to fit the data well. We use variational inference and stochastic optimization to approximate the posterior distribution. The inference procedure scales to over one million cells, whereas competing algorithms do not. Even for smaller datasets, for several tasks, the proposed procedure outperforms state-of-the-art methods like ZIFA and ZINB-WaVE. We also extend our framework to account for batch effects and other confounding factors, and propose a Bayesian hypothesis test for differential expression that outperforms DESeq2.

In this paper we focus our attention on the exploitation of the information contained in financial news to enhance the performance of a classifier of bank distress. Such information should be analyzed and inserted into the predictive model in the most efficient way and this task deals with all the issues related to text analysis and specifically analysis of news media. Among the different models proposed for such purpose, we investigate one of the possible deep learning approaches, based on a doc2vec representation of the textual data, a kind of neural network able to map the sequential and symbolic text input onto a reduced latent semantic space. Afterwards, a second supervised neural network is trained combining news data with standard financial figures to classify banks whether in distressed or tranquil states, based on a small set of known distress events. Then the final aim is not only the improvement of the predictive performance of the classifier but also to assess the importance of news data in the classification process. Does news data really bring more useful information not contained in standard financial variables? Our results seem to confirm such hypothesis.

Multi-task learning is motivated by the observation that humans bring to bear what they know about related problems when solving new ones. Similarly, deep neural networks can profit from related tasks by sharing parameters with other networks. However, humans do not consciously decide to transfer knowledge between tasks. In Natural Language Processing (NLP), it is hard to predict if sharing will lead to improvements, particularly if tasks are only loosely related. To overcome this, we introduce Sluice Networks, a general framework for multi-task learning where trainable parameters control the amount of sharing. Our framework generalizes previous proposals in enabling sharing of all combinations of subspaces, layers, and skip connections. We perform experiments on three task pairs, and across seven different domains, using data from OntoNotes 5.0, and achieve up to 15% average error reductions over common approaches to multi-task learning. We show that a) label entropy is predictive of gains in sluice networks, confirming findings for hard parameter sharing and b) while sluice networks easily fit noise, they are robust across domains in practice.

With our 2018 machine learning predictions, we're taking another shot at machine learning clairvoyance with some brand new calls while also upping the ante to serious "double dog dare you" territory by reiterating some of our previous calls. We'd like to stress that the predictions made here are shared through a lense of "machine learning in the enterprise." As such, we're less concerned with predicting the twists and turns in the heady world of machine learning research and more concerned with the experience of the typical enterprise when looking to leverage the technology to reach its quarterly, annual or longer-term strategic business goals. With that out of the way, let's start by setting the tone based on some recent market research findings on the state of the industry as it stands now. It's a well-known fact that the tech giants have put their dollars where their mouths are when it comes to acquiring machine learning/AI talent.

As mankind expands outwards into the universe, unmanned spacecraft will face a growing problem: as Earth becomes more distant, the transmission time for information and instructions to reach these craft becomes longer and longer. This time lag could make it difficult or even impossible for satellites to respond to fast-moving threats, like space debris, or quickly take opportunities to collect data from unexpected sources, like a passing meteorite. A new grant from NASA to the University of Akron in Ohio will fund research to overcome this issue by helping such spacecraft "think" for themselves using deep-learning artificial intelligence (AI) that works over an Ethereum blockchain network. "I hope to develop technology that can recognize environmental threats and avoid them, as well as complete a number of tasks automatically," Akron Assistant Professor Jin Wei Kocsis, who will lead the research, said in a press release. "I am honored that NASA recognized my work, and I am excited to continue challenging technology's ability to think and do on its own."