Introduction to Linear Regression Analysis

it saids the book is used but like new. However, the book is actually totally new, never used before,no pollution at all. The price is also one of the lowest compare with others. Excellent!!

I have access to this, the third edition and the latest, the fourth edition, through my company's library. There is really no material difference in the content and I was able to save about 80% of the purchase price by buying a used copy of the third edition, vs. new copy of fourth edition. Wonderful book for self study. You will benefit most if you have a good background in probability theory and linear algebra and want to understand the details and language of linear regression. Even without that background chapters one through three will teach you more than you will ever learn in most survey courses in statistics. To fully appreciate the whole book I think you need a one semester course in linear algebra and one or two semesters of probability theory.

This is a good book with good exercises in the end of the chapters, but a little hard to read.

very useful for industrial applications. There are quite a few printing mistakes and that would be a problem for those reader they are not very strong in statistics.

If you have a desire or need to develop regression models, whether for prediction or classification, this is a great place to start climbing the learning curve. The book covers all the essentials, such as how to fit a model to a set of data, how to evaluate the quality of the fit, and how to detect influential data points. It also does a good job with some of the issues involved in fitting a regression (most notably colinearity, overfitting, outliers, and deviations from normality) and discusses ridge regression, principal components regression, and other so-called "robust" methods for dealing with such issues. Even if you plan to use nonlinear modelling techniques like polynomial regression or feed-forward neural networks, this book is worth reading: many of the same issues that are involved when developing linear regression models arise in the context of nonlinear models. I use multivariate polynomial regression models for pricing options, and cite this book in my own recent work on that subject--"Advanced Option Pricing Models" (McGraw Hill, Feb 2005). Jeffrey Owen Katz, Ph.D. Author (with Donna L. McCormick) of "The Encyclopedia of Trading Strategies" (McGraw Hill, 2000).