Researchers in the social sciences are faced with complex data sets in which they have relatively small samples and many variables (high dimensional data). Unlike the various technical guides currently on the market, this book provides and overview of a variety of models alongside clear examples of hands-on application.
Researchers in the social sciences are faced with complex data sets in which they have relatively small samples and many variables (high dimensional data). Unlike the various technical guides currently on the market, Applied Regularization Methods for the Social Sciences provides and overview of a variety of models alongside clear examples of hands-on application. Each chapter in this book covers a specific application of regularization techniques with a user-friendly technical description, followed by examples that provide a thorough demonstration of the methods in action.
Key Features:
- Description of regularization methods in a user friendly and easy to read manner
- Inclusion of regularization-based approaches for a variety of statistical analyses commonly used in the social sciences, including both univariate and multivariate models
- Fully developed extended examples using multiple software packages, including R, SAS, and SPSS
- Website containing all datasets and software scripts used in the examples
- Inclusion of both frequentist and Bayesian regularization approaches
- Application exercises for each chapter that instructors could use in class, and independent researchers could use to practice what they have learned from the book
Arvustused
"The book can be useful to students, instructors, practitioners, and researchers not only in social studies but any areas requiring regularization techniques in application of multivariate statistics to high dimensional data."
Stan Lipovetsky, Minneapolis, USA, Technometrics, August 2022
1. Introduction.
2. Theoretical underpinnings of regularization methods.
3. Regularization methods for linear models.
4. Regularization methods for generalized linear models.
5. Regularization methods for multivariate linear models.
6. Regularization methods for cluster analysis and principal components analysis.
7. Regularization methods for latent variable models.
8. Regularization methods for multilevel models.
9. Advanced topics in feature selection.
Holmes Finch is the George and Frances Ball Distinguished Professor of Educational Psychology at BSU, and a professor of statistics and psychometrics. His research interests include structural equation modeling, item response theory, educational and psychological measurement, multilevel modeling, machine learning, and robust multivariate inference. In addition to conducting research in the field of statistics, he also regularly collaborates with colleagues in fields such as educational psychology, neuropsychology, and exercise physiology.
Lisainfo e-raamatute kohta