Computational Statistics Handbook with MATLAB

Computational Statistics Handbook with MATLAB presents a comprehensive overview of contemporary techniques in computational statistics with a strong emphasis on practical implementation. The authors guide readers through detailed algorithmic descriptions and illustrative examples that illuminate the application of these methods in data analysis. Designed for MATLAB® R2007a and Statistics Toolbox Version 6.0, this handbook builds on established approaches while incorporating innovative computational topics. The text introduces new functions for multivariate normal and multivariate t distributions and provides updated insights into MATLAB’s capabilities for generating univariate and bivariate histograms, glyphs, and parallel coordinate plots. It further explores independent component analysis, nonlinear dimensionality reduction, and multidimensional scaling alongside discussions of linear classifiers, quadratic classifiers, and ensemble methods such as bagging, boosting, and random forests. In addition, the handbook delves into unsupervised learning with a focus on model-based clustering and techniques for assessing clustering outcomes. A dedicated chapter on parametric models covers spline regression, logistic regression, and generalized linear models, while expanded content on smoothers - including bin smoothing, running mean and line smoothers, and smoothing splines - further enriches this essential resource for computational statistics.