We give an overview of recent exciting achievements of deep reinforcement learning (RL). We discuss six core elements, six important mechanisms, and twelve applications. We start with background of machine learning, deep learning and reinforcement learning. Next we discuss core RL elements, including value function, in particular, Deep Q-Network (DQN), policy, reward, model, planning, and exploration. After that, we discuss important mechanisms for RL, including attention and memory, unsupervised learning, transfer learning, multi-agent RL, hierarchical RL, and learning to learn. Then we discuss various applications of RL, including games, in particular, AlphaGo, robotics, natural language processing, including dialogue systems, machine translation, and text generation, computer vision, neural architecture design, business management, finance, healthcare, Industry 4.0, smart grid, intelligent transportation systems, and computer systems. We mention topics not reviewed yet, and list a collection of RL resources. After presenting a brief summary, we close with discussions. 
Please see Deep Reinforcement Learning, arXiv:1810.06339, for a significant update.

Deep Reinforcement Learning: An Overview

In this paper we describe a method to perform sequencediscriminative training of neural network acoustic models without the need for frame-level cross-entropy pre-training. We use the lattice-free version of the maximum mutual information (MMI) criterion: LF-MMI. To make its computation feasible we use a phone n-gram language model, in place of the word language model. To further reduce its space and time complexity we compute the objective function using neural network outputs at one third the standard frame rate. These changes enable us to perform the computation for the forward-backward algorithm on GPUs. Further the reduced output frame-rate also provides a significant speed-up during decoding. We present results on 5 different LVCSR tasks with training data ranging from 100 to 2100 hours. Models trained with LFMMI provide a relative word error rate reduction of ∼11.5%, over those trained with cross-entropy objective function, and ∼8%, over those trained with cross-entropy and sMBR objective functions. A further reduction of ∼2.5%, relative, can be obtained by fine tuning these models with the word-lattice based sMBR objective function.

Purely Sequence-Trained Neural Networks for ASR Based on Lattice-Free MMI.

The most basic assumption used in statistical learning theory is that training data and test data are drawn from the same underlying distribution. Unfortunately, in many applications, the "in-domain" test data is drawn from a distribution that is related, but not identical, to the "out-of-domain" distribution of the training data. We consider the common case in which labeled out-of-domain data is plentiful, but labeled in-domain data is scarce. We introduce a statistical formulation of this problem in terms of a simple mixture model and present an instantiation of this framework to maximum entropy classifiers and their linear chain counterparts. We present efficient inference algorithms for this special case based on the technique of conditional expectation maximization. Our experimental results show that our approach leads to improved performance on three real world tasks on four different data sets from the natural language processing domain.

/pdf/domain-adaptation-for-statistical-classifiers-1owf2of1ps.pdf

Domain adaptation for statistical classifiers

Hidden Markov Models (HMMs) provide a simple and effective framework for modelling time-varying spectral vector sequences. As a consequence, almost all present day large vocabulary continuous speech recognition (LVCSR) systems are based on HMMs.

Whereas the basic principles underlying HMM-based LVCSR are rather straightforward, the approximations and simplifying assumptions involved in a direct implementation of these principles would result in a system which has poor accuracy and unacceptable sensitivity to changes in operating environment. Thus, the practical application of HMMs in modern systems involves considerable sophistication.

The aim of this review is first to present the core architecture of a HMM-based LVCSR system and then describe the various refinements which are needed to achieve state-of-the-art performance. These refinements include feature projection, improved covariance modelling, discriminative parameter estimation, adaptation and normalisation, noise compensation and multi-pass system combination. The review concludes with a case study of LVCSR for Broadcast News and Conversation transcription in order to illustrate the techniques described.

/pdf/application-of-hidden-markov-models-in-speech-recognition-4ic3ad3d5g.pdf

Application of Hidden Markov Models in Speech Recognition

Conversational speech recognition has served as a flagship speech recognition task since the release of the Switchboard corpus in the 1990s. In this paper, we measure the human error rate on the widely used NIST 2000 test set, and find that our latest automated system has reached human parity. The error rate of professional transcribers is 5.9% for the Switchboard portion of the data, in which newly acquainted pairs of people discuss an assigned topic, and 11.3% for the CallHome portion where friends and family members have open-ended conversations. In both cases, our automated system establishes a new state of the art, and edges past the human benchmark, achieving error rates of 5.8% and 11.0%, respectively. The key to our system's performance is the use of various convolutional and LSTM acoustic model architectures, combined with a novel spatial smoothing method and lattice-free MMI acoustic training, multiple recurrent neural network language modeling approaches, and a systematic use of system combination.

/pdf/achieving-human-parity-in-conversational-speech-recognition-31tccanxdu.pdf

Achieving Human Parity in Conversational Speech Recognition

MPE (minimum phone error) is a previously introduced technique for discriminative training of HMM parameters. fMPE applies the same objective function to the features, transforming the data with a kernel-like method and training millions of parameters, comparable to the size of the acoustic model. Despite the large number of parameters, fMPE is robust to over-training. The method is to train a matrix projecting from posteriors of Gaussians to a normal size feature space, and then to add the projected features to normal features such as PLP. The matrix is trained from a zero start using a linear method. Sparsity of posteriors ensures speed in both training and test time. The technique gives similar improvements to MPE (around 10% relative). MPE on top of fMPE results in error rates up to 6.5% relative better than MPE alone, or more if multiple layers of transform are trained.

/pdf/fmpe-discriminatively-trained-features-for-speech-3ltyy4j430.pdf

fMPE: discriminatively trained features for speech recognition

End-to-end learning of recurrent neural networks (RNNs) is an attractive solution for dialog systems; however, current techniques are data-intensive and require thousands of dialogs to learn simple behaviors. We introduce Hybrid Code Networks (HCNs), which combine an RNN with domain-specific knowledge encoded as software and system action templates. Compared to existing end-to-end approaches, HCNs considerably reduce the amount of training data required, while retaining the key benefit of inferring a latent representation of dialog state. In addition, HCNs can be optimized with supervised learning, reinforcement learning, or a mixture of both. HCNs attain state-of-the-art performance on the bAbI dialog dataset (Bordes and Weston, 2016), and outperform two commercially deployed customer-facing dialog systems at our company.

/pdf/hybrid-code-networks-practical-and-efficient-end-to-end-33ti2je43n.pdf

Hybrid Code Networks: practical and efficient end-to-end dialog control with supervised and reinforcement learning

End-to-end learning of recurrent neural networks (RNNs) is an attractive solution for dialog systems; however, current techniques are data-intensive and require thousands of dialogs to learn simple behaviors. We introduce Hybrid Code Networks (HCNs), which combine an RNN with domain-specific knowledge encoded as software and system action templates. Compared to existing end-to-end approaches, HCNs considerably reduce the amount of training data required, while retaining the key benefit of inferring a latent representation of dialog state. In addition, HCNs can be optimized with supervised learning, reinforcement learning, or a mixture of both. HCNs attain state-of-the-art performance on the bAbI dialog dataset, and outperform two commercially deployed customer-facing dialog systems.

This paper describes the technical and system building advances made in IBM's speech recognition technology over the course of the Defense Advanced Research Projects Agency (DARPA) Effective Affordable Reusable Speech-to-Text (EARS) program. At a technical level, these advances include the development of a new form of feature-based minimum phone error training (fMPE), the use of large-scale discriminatively trained full-covariance Gaussian models, the use of septaphone acoustic context in static decoding graphs, and improvements in basic decoding algorithms. At a system building level, the advances include a system architecture based on cross-adaptation and the incorporation of 2100 h of training data in every system component. We present results on English conversational telephony test data from the 2003 and 2004 NIST evaluations. The combination of technical advances and an order of magnitude more training data in 2004 reduced the error rate on the 2003 test set by approximately 21% relative-from 20.4% to 16.1%-over the most accurate system in the 2003 evaluation and produced the most accurate results on the 2004 test sets in every speed category

/pdf/advances-in-speech-transcription-at-ibm-under-the-darpa-ears-4ulo9op63l.pdf

Advances in speech transcription at IBM under the DARPA EARS program

This paper describes the technical advances in IBM's conversational telephony submission to the DARPA-sponsored 2004 rich transcription evaluation (RT-04). These advances include a system architecture based on cross-adaptation; a new form of feature-based MPE training; the use of a full-scale discriminatively trained full covariance Gaussian system; the use of septaphone cross-word acoustic context in static decoding graphs; and the incorporation of 2100 hours of training data in every system component. These advances reduced the error rate by approximately 21% relative, on the 2003 test set, over the best-performing system in last year's evaluation, and produced the best results on the RT-04 current and progress CTS data.

/pdf/the-ibm-2004-conversational-telephony-system-for-rich-5exrlg4l2o.pdf

Geoffrey Zweig

Papers

fMPE: discriminatively trained features for speech recognition

Hybrid Code Networks: practical and efficient end-to-end dialog control with supervised and reinforcement learning

Hybrid Code Networks: practical and efficient end-to-end dialog control with supervised and reinforcement learning

Advances in speech transcription at IBM under the DARPA EARS program

The IBM 2004 conversational telephony system for rich transcription