Recent work in unsupervised feature learning and deep learning has shown that being able to train large models can dramatically improve performance. In this paper, we consider the problem of training a deep network with billions of parameters using tens of thousands of CPU cores. We have developed a software framework called DistBelief that can utilize computing clusters with thousands of machines to train large models. Within this framework, we have developed two algorithms for large-scale distributed training: (i) Downpour SGD, an asynchronous stochastic gradient descent procedure supporting a large number of model replicas, and (ii) Sandblaster, a framework that supports for a variety of distributed batch optimization procedures, including a distributed implementation of L-BFGS. Downpour SGD and Sandblaster L-BFGS both increase the scale and speed of deep network training. We have successfully used our system to train a deep network 100x larger than previously reported in the literature, and achieves state-of-the-art performance on ImageNet, a visual object recognition task with 16 million images and 21k categories. We show that these same techniques dramatically accelerate the training of a more modestly sized deep network for a commercial speech recognition service. Although we focus on and report performance of these methods as applied to training large neural networks, the underlying algorithms are applicable to any gradient-based machine learning algorithm.

Subspace learning seeks a low dimensional representation of data that enables accurate reconstruction. However, in many applications, data is obtained from multiple sources rather than a single source (e.g. an object might be viewed by cameras at different angles, or a document might consist of text and images). The conditional independence of separate sources imposes constraints on their shared latent representation, which, if respected, can improve the quality of the learned low dimensional representation. In this paper, we present a convex formulation of multi-view subspace learning that enforces conditional independence while reducing dimensionality. For this formulation, we develop an efficient algorithm that recovers an optimal data reconstruction by exploiting an implicit convex regularizer, then recovers the corresponding latent representation and reconstruction model, jointly and optimally. Experiments illustrate that the proposed method produces high quality results.

Linear chains and trees are basic building blocks in many applications of graphical models. Although exact inference in these models can be performed by dynamic programming, this computation can still be prohibitively expensive with non-trivial target variable domain sizes due to the quadratic dependence on this size. Standard message-passing algorithms for these problems are inefficient because they compute scores on hypotheses for which there is strong negative local evidence. For this reason there has been significant previous interest in beam search and its variants; however, these methods provide only approximate inference. This paper presents new efficient exact inference algorithms based on the combination of it column generation and pre-computed bounds on the model's cost structure. Improving worst-case performance is impossible. However, our method substantially speeds real-world, typical-case inference in chains and trees. Experiments show our method to be twice as fast as exact Viterbi for Wall Street Journal part-of-speech tagging and over thirteen times faster for a joint part-of-speed and named-entity-recognition task. Our algorithm is also extendable to new techniques for approximate inference, to faster two-best inference, and new opportunities for connections between inference and learning.

In many applications, one has information, e.g., labels that are provided in a semi-supervised manner, about a specific target region of a large data set, and one wants to perform machine learning and data analysis tasks nearby that pre-specified target region. Locally-biased problems of this sort are particularly challenging for popular eigenvector-based machine learning and data analysis tools. At root, the reason is that eigenvectors are inherently global quantities. In this paper, we address this issue by providing a methodology to construct semi-supervised eigenvectors of a graph Laplacian, and we illustrate how these locally-biased eigenvectors can be used to perform locally-biased machine learning. These semi-supervised eigenvectors capture successively-orthogonalized directions of maximum variance, conditioned on being well-correlated with an input seed set of nodes that is assumed to be provided in a semi-supervised manner. We also provide several empirical examples demonstrating how these semi-supervised eigenvectors can be used to perform locally-biased learning.

Sparse learning models typically combine a smooth loss with a nonsmooth penalty, such as trace norm. Although recent developments in sparse approximation have offered promising solution methods, current approaches either apply only to matrix-norm constrained problems or provide suboptimal convergence rates. In this paper, we propose a boosting method for regularized learning that guarantees $\epsilon$ accuracy within $O(1/\epsilon)$ iterations. Performance is further accelerated by interlacing boosting with fixed-rank local optimization---exploiting a simpler local objective than previous work. The proposed method yields state-of-the-art performance on large-scale problems. We also demonstrate an application to latent multiview learning for which we provide the first efficient weak-oracle.

The paper addresses the problem of generating multiple hypotheses for prediction tasks that involve interaction with users or successive components in a cascade. Given a set of multiple hypotheses, such components/users have the ability to automatically rank the results and thus retrieve the best one. The standard approach for handling this scenario is to learn a single model and then produce M-best Maximum a Posteriori (MAP) hypotheses from this model. In contrast, we formulate this multiple {\em choice} learning task as a multiple-output structured-output prediction problem with a loss function that captures the natural setup of the problem. We present a max-margin formulation that minimizes an upper-bound on this loss-function. Experimental results on the problems of image co-segmentation and protein side-chain prediction show that our method outperforms conventional approaches used for this scenario and leads to substantial improvements in prediction accuracy.

Finding maximum aposteriori (MAP) assignments in graphical models is an important task in many applications. Since the problem is generally hard, linear programming (LP) relaxations are often used. Solving these relaxations efficiently is thus an important practical problem. In recent years, several authors have proposed message passing updates corresponding to coordinate descent in the dual LP. However,these are generally not guaranteed to converge to a global optimum. One approach to remedy this is to smooth the LP, and perform coordinate descent on the smoothed dual. However, little is known about the convergence rate of this procedure. Here we perform a thorough rate analysis of such schemes and derive primal and dual convergence rates. We also provide a simple dual to primal mapping that yields feasible primal solutions with a guaranteed rate of convergence. Empirical evaluation supports our theoretical claims and shows that the method is highly competitive with state of the art approaches that yield global optima.

Multi-task sparse feature learning aims to improve the generalization performance by exploiting the shared features among tasks. It has been successfully applied to many applications including computer vision and biomedical informatics. Most of the existing multi-task sparse feature learning algorithms are formulated as a convex sparse regularization problem, which is usually suboptimal, due to its looseness for approximating an $\ell_0$-type regularizer. In this paper, we propose a non-convex formulation for multi-task sparse feature learning based on a novel regularizer. To solve the non-convex optimization problem, we propose a Multi-Stage Multi-Task Feature Learning (MSMTFL) algorithm. Moreover, we present a detailed theoretical analysis showing that MSMTFL achieves a better parameter estimation error bound than the convex formulation. Empirical studies on both synthetic and real-world data sets demonstrate the effectiveness of MSMTFL in comparison with the state of the art multi-task sparse feature learning algorithms.

Diversified ranking is a fundamental task in machine learning. It is broadly applicable inmany real world problems, e.g., information retrieval, team assembling, product search, etc. In this paper, we consider a generic setting where we aim to diversify the top-k ranking list based on an arbitrary relevance function and an arbitrary similarity function among all the examples. We formulate it as an optimization problem and show that in general it is NPhard. Then, we show that for a large volume of the parameter space, the proposed objective function enjoys the diminishing returns property, which enables us to design a scalable, greedy algorithm to find the (1 1/e) near-optimal solution. Experimental results on real data sets demonstrate the effectiveness of the proposed algorithm.

We develop a new algorithm to cluster sparse unweighted graphs -- i.e. partition the nodes into disjoint clusters so that there is higher density within clusters, and low across clusters. By sparsity we mean the setting where both the in-cluster and across cluster edge densities are very small, possibly vanishing in the size of the graph. Sparsity makes the problem noisier, and hence more difficult to solve. Any clustering involves a tradeoff between minimizing two kinds of errors: missing edges within clusters and present edges across clusters. Our insight is that in the sparse case, these must be {\em penalized differently}. We analyze our algorithm's performance on the natural, classical and widely studied ``planted partition'' model (also called the stochastic block model); we show that our algorithm can cluster sparser graphs, and with smaller clusters, than all previous methods. This is seen empirically as well.