arXiv subjects

Shusen Wang

Publications and source records attributed to Shusen Wang.

36 records · Page 2Linked to original sources

arXiv2019

OverSketch: Approximate Matrix Multiplication for the Cloud

We propose OverSketch, an approximate algorithm for distributed matrix multiplication in serverless computing. OverSketch leverages ideas from matrix sketching and high-performance computing to enable cost-efficient multiplication that is resilient to faults and straggling nodes pervasive in low-cost serverless architectures. We establish statistical guarantees on the accuracy of OverSketch and empirically validate our results by solving a large-scale linear program using interior-point methods and demonstrate a 34% reduction in compute time on AWS Lambda.

cs.DC↗

arXiv2019

Scalable Kernel K-Means Clustering with Nystrom Approximation: Relative-Error Bounds

Kernel $k$-means clustering can correctly identify and extract a far more varied collection of cluster structures than the linear $k$-means clustering algorithm. However, kernel $k$-means clustering is computationally expensive when the non-linear feature map is high-dimensional and there are many input points. Kernel approximation, e.g., the Nyström method, has been applied in previous works to approximately solve kernel learning problems when both of the above conditions are present. This work analyzes the application of this paradigm to kernel $k$-means clustering, and shows that applying the linear $k$-means clustering algorithm to $\frac{k}ε (1 + o(1))$ features constructed using a so-called rank-restricted Nyström approximation results in cluster assignments that satisfy a $1 + ε$ approximation ratio in terms of the kernel $k$-means cost function, relative to the guarantee provided by the same algorithm without the use of the Nyström method. As part of the analysis, this work establishes a novel $1 + ε$ relative-error trace norm guarantee for low-rank approximation using the rank-restricted Nyström approximation. Empirical evaluations on the $8.1$ million instance MNIST8M dataset demonstrate the scalability and usefulness of kernel $k$-means clustering with Nyström approximation. This work argues that spectral clustering using Nyström approximation---a popular and computationally efficient, but theoretically unsound approach to non-linear clustering---should be replaced with the efficient and theoretically sound combination of kernel $k$-means clustering with Nyström approximation. The superior performance of the latter approach is empirically verified.

cs.LG↗

arXiv2018

GIANT: Globally Improved Approximate Newton Method for Distributed Optimization

For distributed computing environment, we consider the empirical risk minimization problem and propose a distributed and communication-efficient Newton-type optimization method. At every iteration, each worker locally finds an Approximate NewTon (ANT) direction, which is sent to the main driver. The main driver, then, averages all the ANT directions received from workers to form a {\it Globally Improved ANT} (GIANT) direction. GIANT is highly communication efficient and naturally exploits the trade-offs between local computations and global communications in that more local computations result in fewer overall rounds of communications. Theoretically, we show that GIANT enjoys an improved convergence rate as compared with first-order methods and existing distributed Newton-type methods. Further, and in sharp contrast with many existing distributed Newton-type methods, as well as popular first-order methods, a highly advantageous practical feature of GIANT is that it only involves one tuning parameter. We conduct large-scale experiments on a computer cluster and, empirically, demonstrate the superior performance of GIANT.

cs.LG↗

arXiv2018

Error Estimation for Randomized Least-Squares Algorithms via the Bootstrap

Over the course of the past decade, a variety of randomized algorithms have been proposed for computing approximate least-squares (LS) solutions in large-scale settings. A longstanding practical issue is that, for any given input, the user rarely knows the actual error of an approximate solution (relative to the exact solution). Likewise, it is difficult for the user to know precisely how much computation is needed to achieve the desired error tolerance. Consequently, the user often appeals to worst-case error bounds that tend to offer only qualitative guidance. As a more practical alternative, we propose a bootstrap method to compute a posteriori error estimates for randomized LS algorithms. These estimates permit the user to numerically assess the error of a given solution, and to predict how much work is needed to improve a "preliminary" solution. In addition, we provide theoretical consistency results for the method, which are the first such results in this context (to the best of our knowledge). From a practical standpoint, the method also has considerable flexibility, insofar as it can be applied to several popular sketching algorithms, as well as a variety of error metrics. Moreover, the extra step of error estimation does not add much cost to an underlying sketching algorithm. Finally, we demonstrate the effectiveness of the method with empirical results.

stat.ML↗

arXiv2018

Alchemist: An Apache Spark <=> MPI Interface

The Apache Spark framework for distributed computation is popular in the data analytics community due to its ease of use, but its MapReduce-style programming model can incur significant overheads when performing computations that do not map directly onto this model. One way to mitigate these costs is to off-load computations onto MPI codes. In recent work, we introduced Alchemist, a system for the analysis of large-scale data sets. Alchemist calls MPI-based libraries from within Spark applications, and it has minimal coding, communication, and memory overheads. In particular, Alchemist allows users to retain the productivity benefits of working within the Spark software ecosystem without sacrificing performance efficiency in linear algebra, machine learning, and other related computations. In this paper, we discuss the motivation behind the development of Alchemist, and we provide a detailed overview its design and usage. We also demonstrate the efficiency of our approach on medium-to-large data sets, using some standard linear algebra operations, namely matrix multiplication and the truncated singular value decomposition of a dense matrix, and we compare the performance of Spark with that of Spark+Alchemist. These computations are run on the NERSC supercomputer Cori Phase 1, a Cray XC40.

cs.DC↗

arXiv2018

Accelerating Large-Scale Data Analysis by Offloading to High-Performance Computing Libraries using Alchemist

Apache Spark is a popular system aimed at the analysis of large data sets, but recent studies have shown that certain computations---in particular, many linear algebra computations that are the basis for solving common machine learning problems---are significantly slower in Spark than when done using libraries written in a high-performance computing framework such as the Message-Passing Interface (MPI). To remedy this, we introduce Alchemist, a system designed to call MPI-based libraries from Apache Spark. Using Alchemist with Spark helps accelerate linear algebra, machine learning, and related computations, while still retaining the benefits of working within the Spark environment. We discuss the motivation behind the development of Alchemist, and we provide a brief overview of its design and implementation. We also compare the performances of pure Spark implementations with those of Spark implementations that leverage MPI-based codes via Alchemist. To do so, we use data science case studies: a large-scale application of the conjugate gradient method to solve very large linear systems arising in a speech classification problem, where we see an improvement of an order of magnitude; and the truncated singular value decomposition (SVD) of a 400GB three-dimensional ocean temperature data set, where we see a speedup of up to 7.9x. We also illustrate that the truncated SVD computation is easily scalable to terabyte-sized data by applying it to data sets of sizes up to 17.6TB.

cs.DC↗

arXiv2018

Sketched Ridge Regression: Optimization Perspective, Statistical Perspective, and Model Averaging

We address the statistical and optimization impacts of the classical sketch and Hessian sketch used to approximately solve the Matrix Ridge Regression (MRR) problem. Prior research has quantified the effects of classical sketch on the strictly simpler least squares regression (LSR) problem. We establish that classical sketch has a similar effect upon the optimization properties of MRR as it does on those of LSR: namely, it recovers nearly optimal solutions. By contrast, Hessian sketch does not have this guarantee, instead, the approximation error is governed by a subtle interplay between the "mass" in the responses and the optimal objective value. For both types of approximation, the regularization in the sketched MRR problem results in significantly different statistical properties from those of the sketched LSR problem. In particular, there is a bias-variance trade-off in sketched MRR that is not present in sketched LSR. We provide upper and lower bounds on the bias and variance of sketched MRR, these bounds show that classical sketch significantly increases the variance, while Hessian sketch significantly increases the bias. Empirically, sketched MRR solutions can have risks that are higher by an order-of-magnitude than those of the optimal MRR solutions. We establish theoretically and empirically that model averaging greatly decreases the gap between the risks of the true and sketched solutions to the MRR problem. Thus, in parallel or distributed settings, sketching combined with model averaging is a powerful technique that quickly obtains near-optimal solutions to the MRR problem while greatly mitigating the increased statistical risk incurred by sketching.

stat.ML↗

arXiv2017

Efficient Data-Driven Geologic Feature Detection from Pre-stack Seismic Measurements using Randomized Machine-Learning Algorithm

Conventional seismic techniques for detecting the subsurface geologic features are challenged by limited data coverage, computational inefficiency, and subjective human factors. We developed a novel data-driven geological feature detection approach based on pre-stack seismic measurements. Our detection method employs an efficient and accurate machine-learning detection approach to extract useful subsurface geologic features automatically. Specifically, our method is based on kernel ridge regression model. The conventional kernel ridge regression can be computationally prohibited because of the large volume of seismic measurements. We employ a data reduction technique in combination with the conventional kernel ridge regression method to improve the computational efficiency and reduce memory usage. In particular, we utilize a randomized numerical linear algebra technique, named Nyström method, to effectively reduce the dimensionality of the feature space without compromising the information content required for accurate detection. We provide thorough computational cost analysis to show efficiency of our new geological feature detection methods. We further validate the performance of our new subsurface geologic feature detection method using synthetic surface seismic data for 2D acoustic and elastic velocity models. Our numerical examples demonstrate that our new detection method significantly improves the computational efficiency while maintaining comparable accuracy. Interestingly, we show that our method yields a speed-up ratio on the order of $\sim10^2$ to $\sim 10^3$ in a multi-core computational environment.

cs.LG↗

arXiv2016

Towards More Efficient SPSD Matrix Approximation and CUR Matrix Decomposition

Symmetric positive semi-definite (SPSD) matrix approximation methods have been extensively used to speed up large-scale eigenvalue computation and kernel learning methods. The standard sketch based method, which we call the prototype model, produces relatively accurate approximations, but is inefficient on large square matrices. The Nyström method is highly efficient, but can only achieve low accuracy. In this paper we propose a novel model that we call the {\it fast SPSD matrix approximation model}. The fast model is nearly as efficient as the Nyström method and as accurate as the prototype model. We show that the fast model can potentially solve eigenvalue problems and kernel learning problems in linear time with respect to the matrix size $n$ to achieve $1+ε$ relative-error, whereas both the prototype model and the Nyström method cost at least quadratic time to attain comparable error bound. Empirical comparisons among the prototype model, the Nyström method, and our fast model demonstrate the superiority of the fast model. We also contribute new understandings of the Nyström method. The Nyström method is a special instance of our fast model and is approximation to the prototype model. Our technique can be straightforwardly applied to make the CUR matrix decomposition more efficiently computed without much affecting the accuracy.

cs.LG↗

arXiv2016

SPSD Matrix Approximation vis Column Selection: Theories, Algorithms, and Extensions

Symmetric positive semidefinite (SPSD) matrix approximation is an important problem with applications in kernel methods. However, existing SPSD matrix approximation methods such as the Nyström method only have weak error bounds. In this paper we conduct in-depth studies of an SPSD matrix approximation model and establish strong relative-error bounds. We call it the prototype model for it has more efficient and effective extensions, and some of its extensions have high scalability. Though the prototype model itself is not suitable for large-scale data, it is still useful to study its properties, on which the analysis of its extensions relies. This paper offers novel theoretical analysis, efficient algorithms, and a highly accurate extension. First, we establish a lower error bound for the prototype model and improve the error bound of an existing column selection algorithm to match the lower bound. In this way, we obtain the first optimal column selection algorithm for the prototype model. We also prove that the prototype model is exact under certain conditions. Second, we develop a simple column selection algorithm with a provable error bound. Third, we propose a so-called spectral shifting model to make the approximation more accurate when the eigenvalues of the matrix decay slowly, and the improvement is theoretically quantified. The spectral shifting method can also be applied to improve other SPSD matrix approximation models.

cs.LG↗

arXiv2015

Improved Analyses of the Randomized Power Method and Block Lanczos Method

The power method and block Lanczos method are popular numerical algorithms for computing the truncated singular value decomposition (SVD) and eigenvalue decomposition problems. Especially in the literature of randomized numerical linear algebra, the power method is widely applied to improve the quality of randomized sketching, and relative-error bounds have been well established. Recently, Musco & Musco (2015) proposed a block Krylov subspace method that fully exploits the intermediate results of the power iteration to accelerate convergence. They showed spectral gap-independent bounds which are stronger than the power method by order-of-magnitude. This paper offers novel error analysis techniques and significantly improves the bounds of both the randomized power method and the block Lanczos method. This paper also establishes the first gap-independent bound for the warm-start block Lanczos method.

math.NA↗

arXiv2015

A Practical Guide to Randomized Matrix Computations with MATLAB Implementations

Matrix operations such as matrix inversion, eigenvalue decomposition, singular value decomposition are ubiquitous in real-world applications. Unfortunately, many of these matrix operations so time and memory expensive that they are prohibitive when the scale of data is large. In real-world applications, since the data themselves are noisy, machine-precision matrix operations are not necessary at all, and one can sacrifice a reasonable amount of accuracy for computational efficiency. In recent years, a bunch of randomized algorithms have been devised to make matrix computations more scalable. Mahoney (2011) and Woodruff (2014) have written excellent but very technical reviews of the randomized algorithms. Differently, the focus of this manuscript is on intuition, algorithm derivation, and implementation. This manuscript should be accessible to people with knowledge in elementary matrix algebra but unfamiliar with randomized matrix computations. The algorithms introduced in this manuscript are all summarized in a user-friendly way, and they can be implemented in lines of MATLAB code. The readers can easily follow the implementations even if they do not understand the maths and algorithms.

cs.MS↗

arXiv2015

Adjusting Leverage Scores by Row Weighting: A Practical Approach to Coherent Matrix Completion

Low-rank matrix completion is an important problem with extensive real-world applications. When observations are uniformly sampled from the underlying matrix entries, existing methods all require the matrix to be incoherent. This paper provides the first working method for coherent matrix completion under the standard uniform sampling model. Our approach is based on the weighted nuclear norm minimization idea proposed in several recent work, and our key contribution is a practical method to compute the weighting matrices so that the leverage scores become more uniform after weighting. Under suitable conditions, we are able to derive theoretical results, showing the effectiveness of our approach. Experiments on synthetic data show that our approach recovers highly coherent matrices with high precision, whereas the standard unweighted method fails even on noise-free data.

cs.LG↗

arXiv2014

Sharpened Error Bounds for Random Sampling Based $\ell_2$ Regression

Given a data matrix $X \in R^{n\times d}$ and a response vector $y \in R^{n}$, suppose $n>d$, it costs $O(n d^2)$ time and $O(n d)$ space to solve the least squares regression (LSR) problem. When $n$ and $d$ are both large, exactly solving the LSR problem is very expensive. When $n \gg d$, one feasible approach to speeding up LSR is to randomly embed $y$ and all columns of $X$ into a smaller subspace $R^c$; the induced LSR problem has the same number of columns but much fewer number of rows, and it can be solved in $O(c d^2)$ time and $O(c d)$ space. We discuss in this paper two random sampling based methods for solving LSR more efficiently. Previous work showed that the leverage scores based sampling based LSR achieves $1+ε$ accuracy when $c \geq O(d ε^{-2} \log d)$. In this paper we sharpen this error bound, showing that $c = O(d \log d + d ε^{-1})$ is enough for achieving $1+ε$ accuracy. We also show that when $c \geq O(μd ε^{-2} \log d)$, the uniform sampling based LSR attains a $2+ε$ bound with positive probability.

cs.LG↗

arXiv2014

Efficient Algorithms and Error Analysis for the Modified Nystrom Method

Many kernel methods suffer from high time and space complexities and are thus prohibitive in big-data applications. To tackle the computational challenge, the Nyström method has been extensively used to reduce time and space complexities by sacrificing some accuracy. The Nyström method speedups computation by constructing an approximation of the kernel matrix using only a few columns of the matrix. Recently, a variant of the Nyström method called the modified Nyström method has demonstrated significant improvement over the standard Nyström method in approximation accuracy, both theoretically and empirically. In this paper, we propose two algorithms that make the modified Nyström method practical. First, we devise a simple column selection algorithm with a provable error bound. Our algorithm is more efficient and easier to implement than and nearly as accurate as the state-of-the-art algorithm. Second, with the selected columns at hand, we propose an algorithm that computes the approximation in lower time complexity than the approach in the previous work. Furthermore, we prove that the modified Nyström method is exact under certain conditions, and we establish a lower error bound for the modified Nyström method.

cs.LG↗

arXiv2013

Improving CUR Matrix Decomposition and the Nyström Approximation via Adaptive Sampling

The CUR matrix decomposition and the Nyström approximation are two important low-rank matrix approximation techniques. The Nyström method approximates a symmetric positive semidefinite matrix in terms of a small number of its columns, while CUR approximates an arbitrary data matrix by a small number of its columns and rows. Thus, CUR decomposition can be regarded as an extension of the Nyström approximation. In this paper we establish a more general error bound for the adaptive column/row sampling algorithm, based on which we propose more accurate CUR and Nyström algorithms with expected relative-error bounds. The proposed CUR and Nyström algorithms also have low time complexity and can avoid maintaining the whole data matrix in RAM. In addition, we give theoretical analysis for the lower error bounds of the standard Nyström method and the ensemble Nyström method. The main theoretical results established in this paper are novel, and our analysis makes no special assumption on the data matrices.

cs.LG↗

arXiv2012

A Scalable CUR Matrix Decomposition Algorithm: Lower Time Complexity and Tighter Bound

The CUR matrix decomposition is an important extension of Nyström approximation to a general matrix. It approximates any data matrix in terms of a small number of its columns and rows. In this paper we propose a novel randomized CUR algorithm with an expected relative-error bound. The proposed algorithm has the advantages over the existing relative-error CUR algorithms that it possesses tighter theoretical bound and lower time complexity, and that it can avoid maintaining the whole data matrix in main memory. Finally, experiments on several real-world datasets demonstrate significant improvement over the existing relative-error algorithms.

cs.LG↗

arXiv2012

EP-GIG Priors and Applications in Bayesian Sparse Learning

In this paper we propose a novel framework for the construction of sparsity-inducing priors. In particular, we define such priors as a mixture of exponential power distributions with a generalized inverse Gaussian density (EP-GIG). EP-GIG is a variant of generalized hyperbolic distributions, and the special cases include Gaussian scale mixtures and Laplace scale mixtures. Furthermore, Laplace scale mixtures can subserve a Bayesian framework for sparse learning with nonconvex penalization. The densities of EP-GIG can be explicitly expressed. Moreover, the corresponding posterior distribution also follows a generalized inverse Gaussian distribution. These properties lead us to EM algorithms for Bayesian sparse learning. We show that these algorithms bear an interesting resemblance to iteratively re-weighted $\ell_2$ or $\ell_1$ methods. In addition, we present two extensions for grouped variable selection and logistic regression.

stat.ML↗