welcome

Outlier detection is an important area of machine learning, and I believe this book is a valuable resource to understand the field itself and to understand how to effectively conduct outlier detection projects. It goes through the purposes for outlier detection, the common tools, how they work, their limitations, practical considerations to produce meaningful results, methods to combine detectors, and methods to evaluate results. The book also puts an emphasis on interpretability, allowing users to understand why some items may be considered statistically more unusual than other items.

The book starts with statistical methods, covers traditional machine learning, and finally covers deep learning approaches, applied to tabular data as well as to text, image, network, and time series data. As well as providing a deep background in outlier detection, this provides the reader with the skills to use numerous of the currently available python outlier detection tools, such as those available in scikit-learn and popular packages such as PyOD (PYthon Outlier Detection).