Regression Models in Engineering and the Applied Sciences

This is an introductory book on regression modeling and its applications in various fields.
The aim of the book is to give a clear exposition of the concept of regression and techniques used in computing the unknown parameters of various regression models.

Chapter 1 discusses the historical development of the concept of regression, missing data estimation, regression accuracy metrics, coefficient of determination, censored samples, overfitting and limitations of regression.
Simple linear regression, quantile regression, logistic regression, and count regression models such as discrete Weibull regression, Poisson regression, negative binomial regression, Gaussian and inverse Gaussian regression, Tobit regression, constrained regression, penalized regression, ridge, LASSO, elasticnet, support vector regression, isotonic, change-point, circular, and principal component regression, Cox regression, functional regression, trigonometric and Fourier regression, convex, step-wise, spatial, clustered, Kink, and cointegration regression are presented in Chapter 2.
Classical regression models (simple linear regression, multiple linear regression and polynomial regression) are introduced in Chapter 3.
A novel method to estimate the regression slope without using covariance is described. Data transformation in regression models is discussed at length. A discussion on fitting regression models to frequency-tabulated data appears in Chap. 3.
Chapter 4 discusses logistic and multinomial logistic regression models. A useful discussion on choosing labels to dependent variable appears next. Data transformation for logistic regression is described, as well as a comparison between logit and probit models of logistic regression.