Probability and regression

From MaRDI portal





This book is about the theoretical concepts of regression. Data are almost irrelevant. The first part introduces the basics of probability theory: probability spaces, conditional probability, random variables, expectation, variance and correlation. The second part is on the theory of regression. It starts with conditional expectation. Considered as random variable it is synonymous with regression. The characteristics of regression are discussed thoroughly. Then, simple linear regression, nonlinear, bivariate, conditional, and multivariate regession are dealt with separately. This part has also chapters on conditional variance and covariance and on matrix algebra. The third part discusses the relevance of the concept of regression for applied sciences. The question is if it is possible to describe causal dependencies with this approach. It starts with the presentation of some paradoxa and their solution by randomisation. Individual and mean causal effects are defined and discussed next. The effects are defined as differences of conditional expectations. The last chapter deals with conditional causal effects. Besides the use of a more formal notation, the level of the book is rather elementary. Even simpler concepts are introduced. On the other hand, theoretical concepts are carefully explained. So it may also be useful for applied researchers using regression analysis as an every day tool. Exercises are presented at the end of every chapter. They are rated according to their difficulty. Solutions are given, too.











This page was built for publication: Probability and regression

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q3151806)