Description
Exploratory Data Analysis (EDA) visualizes and summarizes data before making model assumptions to generate hypotheses and deal with large and complex data sets. It seems only natural to apply it to the very popular MATLAB, and the authors do so for practitioners as well as advanced undergraduate or graduate students. They introduce EDA, examine it as a method of pattern discovery in its dimensionality reduction by linear and nonlinear methods, take data tours, investigate clusters and model-based clustering, and describe smoothing subplots. They then describe graphical methods for EDA, including visualizing clusters, and showing distribution shapes and multivariate visualization techniques. Appendices include descriptions of proximity measures, software resources for EDA, data sets, and information on applying MATLAB. They include MATLAB code for nearly all the algorithms in the text and web addresses for the EDA Toolbox. Readers should have a background in linear algebra. Annotation ©2004 Book News, Inc., Portland, OR (booknews.com)
Read More
Exploratory data analysis (EDA) was conceived at a time when computers were not widely used, and thus computational ability was rather limited. As computational sophistication has increased, EDA has become an even more powerful process for visualizing and summarizing data before making model assumptions to generate hypotheses, encompassing larger and more complex data sets. There are many resources for those interested in the theory of EDA, but this is the first book to use MATLAB to illustrate the computational aspects of this discipline.Exploratory Data Analysis with MATLAB presents the methods of EDA from a computational perspective. The authors extensively use MATLAB code and algorithm descriptions to provide state-of-the-art techniques for finding patterns and structure in data. Addressing theory, they also incorporate many annotated references to direct readers to the more theoretical aspects of the methods. The book presents an approach using the basic functions from MATLAB and the MATLAB Statistics Toolbox, in order to be more accessible and enduring. It also contains pseudo-code to enable users of other software packages to implement the algorithms.This text places the tools needed to implement EDA theory at the fingertips of researchers, applied mathematicians, computer scientists, engineers, and statisticians by using a practical/computational approach.
Read Less