In this article a novel approach for training deep neural networks using Bayesian techniques is presented. The Bayesian methodology allows for an easy evaluation of model uncertainty and additionally is robust to overfitting. These are commonly the two main problems classical, i.e. non-Bayesian, architectures have to struggle with. The proposed approach applies variational inference in order to approximate the intractable posterior distribution. In particular, the variational distribution is defined as product of multiple multivariate normal distributions with tridiagonal covariance matrices. Each single normal distribution belongs either to the weights, or to the biases corresponding to one network layer. The layer-wise a posteriori variances are defined based on the corresponding expectation values and further the correlations are assumed to be identical. Therefore, only a few additional parameters need to be optimized compared to non-Bayesian settings. The novel approach is successfully evaluated on basis of the popular benchmark datasets MNIST and CIFAR-10.

Anomaly detection is a classical problem where the aim is to detect anomalous data that do not belong to the normal data distribution. Current state-of-the-art methods for anomaly detection on complex high-dimensional data are based on the generative adversarial network (GAN). However, the traditional GAN loss is not directly aligned with the anomaly detection objective: it encourages the distribution of the generated samples to overlap with the real data and so the resulting discriminator has been found to be ineffective as an anomaly detector. In this paper, we propose simple modifications to the GAN loss such that the generated samples lie at the boundary of the real data distribution. With our modified GAN loss, our anomaly detection method, called Fence GAN (FGAN), directly uses the discriminator score as an anomaly threshold. Our experimental results using the MNIST, CIFAR10 and KDD99 datasets show that Fence GAN yields the best anomaly classification accuracy compared to state-of-the-art methods.

Code-switching, the alternation of languages within a conversation or utterance, is a common communicative phenomenon that occurs in multilingual communities across the world. This survey reviews computational approaches for code-switched Speech and Natural Language Processing. We motivate why processing code-switched text and speech is essential for building intelligent agents and systems that interact with users in multilingual communities. As code-switching data and resources are scarce, we list what is available in various code-switched language pairs with the language processing tasks they can be used for. We review code-switching research in various Speech and NLP applications, including language processing tools and end-to-end systems. We conclude with future directions and open problems in the field.

Recurrent neural networks (RNNs) can model natural language by sequentially "reading" input tokens and outputting a distributed representation of each token. Due to the sequential nature of RNNs, inference time is linearly dependent on the input length, and all inputs are read regardless of their importance. Efforts to speed up this inference, known as "neural speed reading", either ignore or skim over part of the input. We present Structural-Jump-LSTM: the first neural speed reading model to both skip and jump text during inference. The model consists of a standard LSTM and two agents: one capable of skipping single words when reading, and one capable of exploiting punctuation structure (sub-sentence separators (,:), sentence end symbols (.!?), or end of text markers) to jump ahead after reading a word. A comprehensive experimental evaluation of our model against all five state-of-the-art neural reading models shows that Structural-Jump-LSTM achieves the best overall floating point operations (FLOP) reduction (hence is faster), while keeping the same accuracy or even improving it compared to a vanilla LSTM that reads the whole text. Recurrent neural networks (RNNs) are a popular model for processing sequential data.

Prognostics or Remaining Useful Life (RUL) Estimation from multi-sensor time series data is useful to enable condition-based maintenance and ensure high operational availability of equipment. We propose a novel deep learning based approach for Prognostics with Uncertainty Quantification that is useful in scenarios where: (i) access to labeled failure data is scarce due to rarity of failures (ii) future operational conditions are unobserved and (iii) inherent noise is present in the sensor readings. All three scenarios mentioned are unavoidable sources of uncertainty in the RUL estimation process often resulting in unreliable RUL estimates. To address (i), we formulate RUL estimation as an Ordinal Regression (OR) problem, and propose LSTM-OR: deep Long Short Term Memory (LSTM) network based approach to learn the OR function. We show that LSTM-OR naturally allows for incorporation of censored operational instances in training along with the failed instances, leading to more robust learning. To address (ii), we propose a simple yet effective approach to quantify predictive uncertainty in the RUL estimation models by training an ensemble of LSTM-OR models. Through empirical evaluation on C-MAPSS turbofan engine benchmark datasets, we demonstrate that LSTM-OR is significantly better than the commonly used deep metric regression based approaches for RUL estimation, especially when failed training instances are scarce. Further, our uncertainty quantification approach yields high quality predictive uncertainty estimates while also leading to improved RUL estimates compared to single best LSTM-OR models.

In this work, we present a method for node embedding in temporal graphs. We propose an algorithm that learns the evolution of a temporal graph's nodes and edges over time and incorporates this dynamics in a temporal node embedding framework for different graph prediction tasks. We present a joint loss function that creates a temporal embedding of a node by learning to combine its historical temporal embeddings, such that it optimizes per given task (e.g., link prediction). The algorithm is initialized using static node embeddings, which are then aligned over the representations of a node at different time points, and eventually adapted for the given task in a joint optimization. We evaluate the effectiveness of our approach over a variety of temporal graphs for the two fundamental tasks of temporal link prediction and multi-label node classification, comparing to competitive baselines and algorithmic alternatives. Our algorithm shows performance improvements across many of the datasets and baselines and is found particularly effective for graphs that are less cohesive, with a lower clustering coefficient.

Modern deep learning models involve a huge number of parameters. In nearly all applications of these models, current practice suggests that we should design the network to be sufficiently complex so that the model (as trained, typically, by gradient descent) interpolates the data, i.e., achieves zero training error. Indeed, in a thought-provoking experiment, Zhang et al. (2016) showed that state-of-the-art deep neural network architectures can be trained to interpolate the data even when the actual labels are replaced by entirely random ones. Despite their enormous complexity, deep neural networks are frequently seen to generalize well, in meaningful practical problems. At first sight, this seems to defy conventional statistical wisdom: interpolation (vanishing training error) is usually taken to be a proxy for overfitting or poor generalization (large gap between training and test error). In an insightful series of papers, Belkin et al. (2018b,c,a) pointed out that these concepts are, in general, distinct, and interpolation does not contradict generalization. For example, kernel ridge regression is a relatively well-understood setting in which interpolation can coexist with good generalization (Liang and Rakhlin, 2018). In this paper, we examine the prediction risk of minimum l norm or "ridgeless" least squares regression, under

Though its effectiveness has been demonstrated in MNIST and smallNORB datasets, the networks still face challenges in other datasets for images with distinct contexts. In this research, we improve the design of CN (Vector version) namely we expand more Pooling layers to filter image backgrounds and increase Reconstruction layers to make better image restoration. Additionally, we perform experiments to compare accuracy and speed of CN versus DL models. In DL models, we utilize Inception V3 and DenseNet V201 for powerful computers besides NASNet, MobileNet V1 and MobileNet V2 for small and embedded devices. We evaluate our models on a fingerspelling alphabet dataset from American Sign Language (ASL). The results show that CNs perform comparably to DL models while dramatically reducing training time. We also make a demonstration and give a link for the purpose of illustration.

Most deep learning-based models for speech enhancement have mainly focused on estimating the magnitude of spectrogram while reusing the phase from noisy speech for reconstruction. This is due to the difficulty of estimating the phase of clean speech. To improve speech enhancement performance, we tackle the phase estimation problem in three ways. First, we propose Deep Complex U-Net, an advanced U-Net structured model incorporating well-defined complex-valued building blocks to deal with complex-valued spectrograms. Second, we propose a polar coordinate-wise complex-valued masking method to reflect the distribution of complex ideal ratio masks. Third, we define a novel loss function, weighted source-to-distortion ratio (wSDR) loss, which is designed to directly correlate with a quantitative evaluation measure. Our model was evaluated on a mixture of the Voice Bank corpus and DEMAND database, which has been widely used by many deep learning models for speech enhancement. Ablation experiments were conducted on the mixed dataset showing that all three proposed approaches are empirically valid. Experimental results show that the proposed method achieves state-of-the-art performance in all metrics, outperforming previous approaches by a large margin.

We propose a novel learning paradigm for Deep Neural Networks (DNN) by using Boolean logic algebra. We first present the basic differentiable operators of a Boolean system such as conjunction, disjunction and exclusive-OR and show how these elementary operators can be combined in a simple and meaningful way to form Neural Logic Networks (NLNs). We examine the effectiveness of the proposed NLN framework in learning Boolean functions and discrete-algorithmic tasks. We demonstrate that, in contrast to the implicit learning in MLP approach, the proposed neural logic networks can learn the logical functions explicitly that can be verified and interpreted by human. In particular, we propose a new framework for learning the inductive logic programming (ILP) problems by exploiting the explicit representational power of NLN. We show the proposed neural ILP solver is capable of feats such as predicate invention and recursion and can outperform the current state of the art neural ILP solvers using a variety of benchmark tasks such as decimal addition and multiplication, and sorting on ordered list.