Kontoghiorghes E.J. (ed.) Handbook of Parallel Computing and Statistics

Chapman & Hall/CRC, 2006, -510 p.Over the last decades a wealth of parallel algorithms has been discovered for solving a wide range of problems that arise in diverse applications areas. Effort has been concentrated mainly on the solution of large scale industrial and engineering problems. Although some of these application areas, such as signal processing and pattern recognition, involve significant statistical computing components, the development of parallel algorithms for general use in statistics and econometrics remains comparatively neglected. This is, to some extent, due to a lack of strong interaction between the parallel computing and statistical communities.
A number of current parallel numerical libraries provide subroutines that are useful to statisticians. For example, most of them offer routines to solve constrained least squares problems and matrix problems that arise in statistical modeling and estimation. However, these routines have been constructed as numerical tools for general use and are unsuitable for the efficient solution of statistical problems that exhibit special properties and characteristics. The design of specifically targeted parallel numerical libraries and tools to facilitate the solution of computationally intensive statistical problems requires close collaboration between statisticians and parallel computing experts.
The aim of this handbook is twofold: first, to provide an overview of the state-of-the-art in parallel algorithms and processing from a statistical computing standpoint; and, second, to contribute toward the development and deepening of research in the interface between parallel and statistical computation.General — Parallel Computing
A Brief Introduction to Parallel Computing
Parallel Computer Architecture
Fortran and Java for High-Performance Computing
Parallel Algorithms for the Singular Value Decomposition
Iterative Methods for the Partial Eigensolution of Symmetric Matrices on Parallel Machines
Optimization
Parallel Optimization Methods
Parallel Computing in Global Optimization
Nonlinear Optimization: A Parallel Linear Algebra Standpoint
Statistical Applications
On Some Statistical Methods for Parallel Computation
Parallel Algorithms for PredictiveModeling
Parallel Programs for Adaptive Designs
A Modular VLSI Architecture for the Real-Time Estimation of Higher Order Moments and Cumulants
Principal Component Analysis for Information Retrieval
Matrix Rank Reduction for Data Analysis and Feature Extraction
Parallel Computation in Econometrics: A Simplified Approach
Parallel Bayesian Computation