Deep learning is a form of machine learning that enables computers to learn from experience and understand the world in terms of a hierarchy of concepts. Because the computer gathers knowledge from experience, there is no need for a human computer operator to formally specify all the knowledge that the computer needs. The hierarchy of concepts allows the computer to learn complicated concepts by building them out of simpler ones; a graph of these hierarchies would be many layers deep. This book introduces a broad range of topics in deep learning. The text offers mathematical and conceptual background, covering relevant concepts in linear algebra, probability theory and information theory, numerical computation, and machine learning. It describes deep learning techniques used by practitioners in industry, including deep feedforward networks, regularization, optimization algorithms, convolutional networks, sequence modeling, and practical methodology; and it surveys such applications as natural language processing, speech recognition, computer vision, online recommendation systems, bioinformatics, and videogames. Finally, the book offers research perspectives, covering such theoretical topics as linear factor models, autoencoders, representation learning, structured probabilistic models, Monte Carlo methods, the partition function, approximate inference, and deep generative models. Deep Learning can be used by undergraduate or graduate students planning careers in either industry or research, and by software engineers who want to begin using deep learning in their products or platforms. A website offers supplementary material for both readers and instructors.

Deep Learning

Recent methods for learning vector space representations of words have succeeded in capturing fine-grained semantic and syntactic regularities using vector arithmetic, but the origin of these regularities has remained opaque. We analyze and make explicit the model properties needed for such regularities to emerge in word vectors. The result is a new global logbilinear regression model that combines the advantages of the two major model families in the literature: global matrix factorization and local context window methods. Our model efficiently leverages statistical information by training only on the nonzero elements in a word-word cooccurrence matrix, rather than on the entire sparse matrix or on individual context windows in a large corpus. The model produces a vector space with meaningful substructure, as evidenced by its performance of 75% on a recent word analogy task. It also outperforms related models on similarity tasks and named entity recognition.

/pdf/glove-global-vectors-for-word-representation-55l84hyy3b.pdf

Glove: Global Vectors for Word Representation

We propose two novel model architectures for computing continuous vector representations of words from very large data sets. The quality of these representations is measured in a word similarity task, and the results are compared to the previously best performing techniques based on different types of neural networks. We observe large improvements in accuracy at much lower computational cost, i.e. it takes less than a day to learn high quality word vectors from a 1.6 billion words data set. Furthermore, we show that these vectors provide state-of-the-art performance on our test set for measuring syntactic and semantic word similarities.

Efficient Estimation of Word Representations in Vector Space

In this paper, we propose a novel neural network model called RNN Encoder‐ Decoder that consists of two recurrent neural networks (RNN). One RNN encodes a sequence of symbols into a fixedlength vector representation, and the other decodes the representation into another sequence of symbols. The encoder and decoder of the proposed model are jointly trained to maximize the conditional probability of a target sequence given a source sequence. The performance of a statistical machine translation system is empirically found to improve by using the conditional probabilities of phrase pairs computed by the RNN Encoder‐Decoder as an additional feature in the existing log-linear model. Qualitatively, we show that the proposed model learns a semantically and syntactically meaningful representation of linguistic phrases.

/pdf/learning-phrase-representations-using-rnn-encoder-decoder-74734oka8k.pdf

Learning Phrase Representations using RNN Encoder--Decoder for Statistical Machine Translation

The success of machine learning algorithms generally depends on data representation, and we hypothesize that this is because different representations can entangle and hide more or less the different explanatory factors of variation behind the data. Although specific domain knowledge can be used to help design representations, learning with generic priors can also be used, and the quest for AI is motivating the design of more powerful representation-learning algorithms implementing such priors. This paper reviews recent work in the area of unsupervised feature learning and deep learning, covering advances in probabilistic models, autoencoders, manifold learning, and deep networks. This motivates longer term unanswered questions about the appropriate objectives for learning good representations, for computing representations (i.e., inference), and the geometrical connections between representation learning, density estimation, and manifold learning.

/pdf/representation-learning-a-review-and-new-perspectives-4sidphwbhg.pdf

Representation Learning: A Review and New Perspectives

We introduce a novel machine learning framework based on recursive autoencoders for sentence-level prediction of sentiment label distributions. Our method learns vector space representations for multi-word phrases. In sentiment prediction tasks these representations outperform other state-of-the-art approaches on commonly used datasets, such as movie reviews, without using any pre-defined sentiment lexica or polarity shifting rules. We also evaluate the model's ability to predict sentiment distributions on a new dataset based on confessions from the experience project. The dataset consists of personal user stories annotated with multiple labels which, when aggregated, form a multinomial distribution that captures emotional reactions. Our algorithm can more accurately predict distributions over such labels compared to several competitive baselines.

/pdf/semi-supervised-recursive-autoencoders-for-predicting-3oakyipcgn.pdf

Semi-Supervised Recursive Autoencoders for Predicting Sentiment Distributions

Unsupervised word representations are very useful in NLP tasks both as inputs to learning algorithms and as extra word features in NLP systems. However, most of these models are built with only local context and one representation per word. This is problematic because words are often polysemous and global context can also provide useful information for learning word meanings. We present a new neural network architecture which 1) learns word embeddings that better capture the semantics of words by incorporating both local and global document context, and 2) accounts for homonymy and polysemy by learning multiple embeddings per word. We introduce a new dataset with human judgments on pairs of words in sentential context, and evaluate our model on it, showing that our model outperforms competitive baselines and other neural language models.

/pdf/improving-word-representations-via-global-context-and-4x4fd6wohu.pdf

Improving Word Representations via Global Context and Multiple Word Prototypes

Paraphrase detection is the task of examining two sentences and determining whether they have the same meaning. In order to obtain high accuracy on this task, thorough syntactic and semantic analysis of the two statements is needed. We introduce a method for paraphrase detection based on recursive autoencoders (RAE). Our unsupervised RAEs are based on a novel unfolding objective and learn feature vectors for phrases in syntactic trees. These features are used to measure the word- and phrase-wise similarity between two sentences. Since sentences may be of arbitrary length, the resulting matrix of similarity measures is of variable size. We introduce a novel dynamic pooling layer which computes a fixed-sized representation from the variable-sized matrices. The pooled representation is then used as input to a classifier. Our method outperforms other state-of-the-art approaches on the challenging MSRP paraphrase corpus.

/pdf/dynamic-pooling-and-unfolding-recursive-autoencoders-for-2yy0n1z72y.pdf

Dynamic Pooling and Unfolding Recursive Autoencoders for Paraphrase Detection

A central challenge in human computation is in understanding how to design task environments that effectively attract participants and coordinate the problem solving process. In this paper, we consider a common problem that requesters face on Amazon Mechanical Turk: how should a task be designed so as to induce good output from workers? In posting a task, a requester decides how to break down the task into unit tasks, how much to pay for each unit task, and how many workers to assign to a unit task. These design decisions affect the rate at which workers complete unit tasks, as well as the quality of the work that results. Using image labeling as an example task, we consider the problem of designing the task to maximize the number of quality tags received within given time and budget constraints. We consider two different measures of work quality, and construct models for predicting the rate and quality of work based on observations of output to various designs. Preliminary results show that simple models can accurately predict the quality of output per unit task, but are less accurate in predicting the rate at which unit tasks complete. At a fixed rate of pay, our models generate different designs depending on the quality metric, and optimized designs obtain significantly more quality tags than baseline comparisons.

/pdf/toward-automatic-task-design-a-progress-report-11p623634p.pdf

Toward automatic task design: a progress report

We consider the possible existence of a manipulator in a prediction market, whose incentive is to maximally increase the predicted probability of an event, and for whom profit or loss in the market is immaterial. We characterize the equilibria in a single-round market scoring rule (MSR), showing that the manipulator will play a strategy that mixes between pretending to have received one of the top signals. We propose a modification to the MSR in the form of trade limits, a maximum amount by which the price of the security can change at a given round. We show analytically that without a manipulator, this process converges to the true posterior, and computationally that in a market with a manipulator, the limits help reduce the distortion by the manipulator when traders do not know about the manipulator's existence. Specifically, we show through simulations that with high probability the honest traders will fully reveal their signals before the manipulator does, and that the price at this point of full revelation by the honest traders can be a significantly better approximation of the true posterior than the ultimate price reached, suggesting a rule by which the market should be stopped at that point.

/pdf/price-manipulation-in-prediction-markets-analysis-and-1wzolag2bx.pdf

Eric H. Huang

Papers

Semi-Supervised Recursive Autoencoders for Predicting Sentiment Distributions

Improving Word Representations via Global Context and Multiple Word Prototypes

Dynamic Pooling and Unfolding Recursive Autoencoders for Paraphrase Detection

Toward automatic task design: a progress report

Price manipulation in prediction markets: analysis and mitigation