Data Analysis and Data Mining An Introduction

by Azzalini, Adelchi; Scarpa, Bruno

• ISBN13: 9780199767106
• ISBN10: 0199767106
• eBook ISBN(s): 9780199942718
• Format: Hardcover
• Copyright: 2012-04-23
• Publisher: Oxford University Press
• More Book Details

Summary

An introduction to statistical data mining,Data Analysis and Data Miningis both textbook and professional resource. Assuming only a basic knowledge of statistical reasoning, it presents core concepts in data mining and exploratory statistical models to students and professional statisticians-both those working in communications and those working in a technological or scientific capacity-who have a limited knowledge of data mining. This book presents key statistical concepts by way of case studies, giving readers the benefit of learning from real problems and real data. Aided by a diverse range of statistical methods and techniques, readers will move from simple problems to complex problems. Through these case studies, authors Adelchi Azzalini and Bruno Scarpa explain exactly how statistical methods work; rather than relying on the "push the button" philosophy, they demonstrate how to use statistical tools to find the best solution to any given problem. Case studies feature current topics highly relevant to data mining, such web page traffic; the segmentation of customers; selection of customers for direct mail commercial campaigns; fraud detection; and measurements of customer satisfaction. Appropriate for both advanced undergraduate and graduate students, this much-needed book will fill a gap between higher level books, which emphasize technical explanations, and lower level books, which assume no prior knowledge and do not explain the methodology behind the statistical operations.

Author Biography

Adelchi Azzalini is Professor of Statistics in the Department of Statistical Sciences at the University of Padua. He actively researches in statistical methodology and has maintained an interest in statistical applications. Bruno Scarpa is Researcher in the Department of Statistical Sciences at the University of Padua. He previously worked as a statistician and data miner in marketing divisions of insurance and telecommunication companies.

Table of Contents

Preface	p. vii
Preface to the English Edition	p. ix
Introduction	p. 1
New problems and new opportunities	p. 1
All models are wrong	p. 9
A matter of style	p. 12
A-B-C	p. 15
Old friends: Linear models	p. 15
Computational aspects	p. 30
Likelihood	p. 33
Logistic regression and GLM	p. 40
Exercises	p. 44
Optimism, Conflicts, and Trade-offs	p. 45
Matching the conceptual frame and real life	p. 45
A simple prototype problem	p. 46
If we knew f(x)…	p. 47
But as we do not know f(x)…	p. 51
Methods for model selection	p. 52
Reduction of dimensions and selection of most appropriate model	p. 58
Exercises	p. 66
Prediction of Quantitative Variables	p. 68
Nonparametric estimation: Why?	p. 68
Local regression	p. 69
The curse of dimensionality	p. 78
Splines	p. 79
Additive models and GAM	p. 89
Projection pursuit	p. 93
Inferential aspects	p. 94
Regression trees	p. 98
Neural networks	p. 106
Case studies	p. 111
Exercises	p. 132
Methods of Classification	p. 134
Prediction of categorical variables	p. 134
An introduction based on a marketing problem	p. 135
Extension to several categories	p. 142
Classification via linear regression	p. 149
Discriminant analysis	p. 154
Some nonparametric methods	p. 159
Classification trees	p. 164
Some other topics	p. 168
Combination of classifiers	p. 176
Case studies	p. 183
Exercises	p. 210
Methods of Internal Analysis	p. 212
Cluster analysis	p. 212
Associations among variables	p. 222
Case study: Web usage mining	p. 232
Complements of Mathematics and Statistics	p. 240
Concepts on linear algebra	p. 240
Concepts of probability theory	p. 241
Concepts of linear models	p. 246
Data Sets	p. 254
Simulated data	p. 254
Car data	p. 254
Brazilian bank data	p. 255
Data for telephone company customers	p. 256
Insurance data	p. 257
Choice of fruit juice data	p. 258
Customer satisfaction	p. 259
Web usage data	p. 261
Symbols and Acronyms	p. 263
References	p. 265
Author Index	p. 269
Subject Index	p. 271
Table of Contents provided by Ingram. All Rights Reserved.