"The purpose of this book is to lay the groundwork for robust data science using rankbased methods. The field of machine learning has not yet fully embraced a class of robust estimators that would address issues that limit the value of least-squares estimation. For example, outliers in data sets may produce misleading results that are not suitable for inference. They can also affect results obtained from penalty estimators. We believe that robust estimators for regression problems are well-suited to data science. This book ...
Read More
"The purpose of this book is to lay the groundwork for robust data science using rankbased methods. The field of machine learning has not yet fully embraced a class of robust estimators that would address issues that limit the value of least-squares estimation. For example, outliers in data sets may produce misleading results that are not suitable for inference. They can also affect results obtained from penalty estimators. We believe that robust estimators for regression problems are well-suited to data science. This book is intended to provide both practical and mathematical foundations in the study of rank-based methods. It will introduce a number of new ideas and approaches to the practice and theory of robust estimation and encourage readers to pursue further investigation in this field. While the main goal of this book is to provide a rigorous treatment of the subject matter, we begin with some introductory material to build insight and intuition about rank-based regression and penalty estimators, especially for those who are new to the topic and those looking to understand key concepts. To motivate the need for such methods, we will start with a discussion of the median as it is the key to rank-based methods and then build on that concept towards the notion of robust data science"--
Read Less
