Computational Statistics Handbook with MATLAB (Chapman & Hall/CRC Computer Science & Data Analysis)

Approaching computational statistics through its theoretical aspects can be daunting. Often intimidated or distracted by the theory, researchers and students can lose sight of the actual goals and applications of the subject. What they need are its key concepts, an understanding of its methods, experience with its implementation, and practice with computational software. Focusing on the computational aspects of statistics rather than the theoretical, Computational Statistics Handbook with MATLAB uses a down-to-earth approach that makes statistics accessible to a wide range of users. The authors integrate the use of MATLAB throughout the book, allowing readers to see the actual implementation of algorithms, but also include step-by-step procedures to allow implementation with any suitable software. The book concentrates on the simulation/Monte Carlo point of view, and contains algorithms for exploratory data analysis, modeling, Monte Carlo simulation, pattern recognition, bootstrap, classification, cross-validation methods, probability density estimation, random number generation, and other computational statistics methods. Emphasis on the practical aspects of statistics, details of the latest techniques, and real implementation experience make the Computational Statistics Handbook with MATLAB more than just the first book to use MATLAB to solve computational problems in statistics. It also forms an outstanding, introduction to statistics for anyone in the many disciplines that involve data analysis. Computational statistics is a fascinating and relatively new field within statistics. While much of classical statistics relies on parameterized functions and related assumptions, the computational statistics approach is to let the data tell the story. The advent of computers with their number-crunching capability, as well as their power to show on the screen two- and three-dimensional structures, has made computational statistics available for any data analyst to use. Computational statistics has a lot to offer the researcher faced with a file full of numbers. The methods of computational statistics can provide assistance ranging from preliminary exploratory data analysis to sophisticated probability density estimation techniques, Monte Carlo methods, and powerful multi-dimensional visualization. All of this power and novel ways of looking at data are accessible to researchers in their daily data analysis tasks. One purpose of this book is to facilitate the exploration of these methods and approaches and to provide the tools to make of this, not just a theoretical exploration, but a practical one. This handbook uses a down-to-earth approach that makes statistics accessible to a wide range of users. The handbook includes MATLAB functions for common probability distributions and features MATLAB interfaces for algorithms and demos Obviously, computational statistics relates to the traditional discipline of statistics.