We trained a large, deep convolutional neural network to classify the 1.2 million high-resolution images in the ImageNet LSVRC-2010 contest into the 1000 different classes. On the test data, we achieved top-1 and top-5 error rates of 37.5% and 17.0% which is considerably better than the previous state-of-the-art. The neural network, which has 60 million parameters and 650,000 neurons, consists of five convolutional layers, some of which are followed by max-pooling layers, and three fully-connected layers with a final 1000-way softmax. To make training faster, we used non-saturating neurons and a very efficient GPU implementation of the convolution operation. To reduce overriding in the fully-connected layers we employed a recently-developed regularization method called "dropout" that proved to be very effective. We also entered a variant of this model in the ILSVRC-2012 competition and achieved a winning top-5 test error rate of 15.3%, compared to 26.2% achieved by the second-best entry.

/pdf/imagenet-classification-with-deep-convolutional-neural-1190x9a6b9.pdf

ImageNet Classification with Deep Convolutional Neural Networks

Multilayer neural networks trained with the back-propagation algorithm constitute the best example of a successful gradient based learning technique. Given an appropriate network architecture, gradient-based learning algorithms can be used to synthesize a complex decision surface that can classify high-dimensional patterns, such as handwritten characters, with minimal preprocessing. This paper reviews various methods applied to handwritten character recognition and compares them on a standard handwritten digit recognition task. Convolutional neural networks, which are specifically designed to deal with the variability of 2D shapes, are shown to outperform all other techniques. Real-life document recognition systems are composed of multiple modules including field extraction, segmentation recognition, and language modeling. A new learning paradigm, called graph transformer networks (GTN), allows such multimodule systems to be trained globally using gradient-based methods so as to minimize an overall performance measure. Two systems for online handwriting recognition are described. Experiments demonstrate the advantage of global training, and the flexibility of graph transformer networks. A graph transformer network for reading a bank cheque is also described. It uses convolutional neural network character recognizers combined with global training techniques to provide record accuracy on business and personal cheques. It is deployed commercially and reads several million cheques per day.

Gradient-based learning applied to document recognition

LIBSVM is a library for Support Vector Machines (SVMs). We have been actively developing this package since the year 2000. The goal is to help users to easily apply SVM to their applications. LIBSVM has gained wide popularity in machine learning and many other areas. In this article, we present all implementation details of LIBSVM. Issues such as solving SVM optimization problems theoretical convergence multiclass classification probability estimates and parameter selection are discussed in detail.

/pdf/libsvm-a-library-for-support-vector-machines-2gdxasg9zv.pdf

LIBSVM: A library for support vector machines

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

Deep neural nets with a large number of parameters are very powerful machine learning systems. However, overfitting is a serious problem in such networks. Large networks are also slow to use, making it difficult to deal with overfitting by combining the predictions of many different large neural nets at test time. Dropout is a technique for addressing this problem. The key idea is to randomly drop units (along with their connections) from the neural network during training. This prevents units from co-adapting too much. During training, dropout samples from an exponential number of different "thinned" networks. At test time, it is easy to approximate the effect of averaging the predictions of all these thinned networks by simply using a single unthinned network that has smaller weights. This significantly reduces overfitting and gives major improvements over other regularization methods. We show that dropout improves the performance of neural networks on supervised learning tasks in vision, speech recognition, document classification and computational biology, obtaining state-of-the-art results on many benchmark data sets.

/pdf/dropout-a-simple-way-to-prevent-neural-networks-from-4gtr92okhv.pdf

Dropout: a simple way to prevent neural networks from overfitting

Systems and methods are disclosed that facilitate producing probabilistic outputs also referred to as posterior probabilities. The probabilistic outputs include an estimate of classification strength. The present invention intercepts non-probabilistic classifier output and applies a set of kernel models based on a softmax function to derive the desired probabilistic outputs. Such probabilistic outputs can be employed with handwriting recognition where the probability of a handwriting sample classification is combined with language models to make better classification decisions.

Probability estimate for K-nearest neighbor

The locally linear embedding (LLE) is improved by introducing multiple linearly independent local weight vectors for each neighborhood. We characterize the reconstruction weights and show the existence of the linearly independent weight vectors at each neighborhood. The modified locally linear embedding (MLLE) proposed in this paper is much stable. It can retrieve the ideal embedding if MLLE is applied on data points sampled from an isometric manifold. MLLE is also compared with the local tangent space alignment (LTSA). Numerical examples are given that show the improvement and efficiency of MLLE.

MLLE: Modified Locally Linear Embedding Using Multiple Weights

A regression-based residual echo suppression (RES) system and process for suppressing the portion of the microphone signal corresponding to a playback of a speaker audio signal that was not suppressed by an acoustic echo canceller (AEC). In general, a prescribed regression technique is used between a prescribed spectral attribute of multiple past and present, fixed-length, periods (e.g., frames) of the speaker signal and the same spectral attribute of a current period (e.g., frame) of the echo residual in the output of the AEC. This automatically takes into consideration the correlation between the time periods of the speaker signal. The parameters of the regression can be easily tracked using adaptive methods. Multiple applications of RES can be used to produce better results and this system and process can be applied to stereo-RES as well.

System and process for regression-based residual acoustic echo suppression

Discriminative learning methods for classification perform well when training and test data are drawn from the same distribution In many situations, though, we have labeled training data for a source domain, and we wish to learn a classifier which performs well on a target domain with a different distribution Under what conditions can we adapt a classifier trained on the source domain for use in the target domain? Intuitively, a good feature representation is a crucial factor in the success of domain adaptation We formalize this intuition theoretically with a generalization bound for domain adaption Our theory illustrates the tradeoffs inherent in designing a representation for domain adaptation and gives a new justification for a recently proposed model It also points toward a promising new model for domain adaptation: one which explicitly minimizes the difference between the source and target domains, while at the same time maximizing the margin of the training set

Analysis of Representations for Domain Adaptation

The described implementations relate to processing of electronic data. One implementation is manifest as a technique that can include obtaining a relational query that references one or more data items and associating progress intervals with the data items. The technique can also include converting the relational query into a corresponding streaming query, and providing the streaming query and the data items with the progress intervals to a stream engine that produces incremental results of the query. For example, the progress intervals can be based on row numbers of a relational database table. The progress intervals can be used to define event lifetimes of streaming events that are provided as inputs to the stream engine.

John Platt

Papers

Probability estimate for K-nearest neighbor

MLLE: Modified Locally Linear Embedding Using Multiple Weights

System and process for regression-based residual acoustic echo suppression

Analysis of Representations for Domain Adaptation

Progressive query computation using streaming architectures