Robust statistics is at the forefront of statistical research, and a central topic in multidisciplinary science where mathematical ideas are used to model and understand the real world, without being affected by contamination that could occur in the data. Nowadays, with the increasing availability of Big data, robust statistical methods are crucially needed. The aim of this course is three-fold: present novel models to describe real-world phenomena, and their properties; study their behavior under different kinds of contamination; discuss their implementation through computational methods. The lectures are addressed to Master, PhD students and postdocs.
Topics: International School on Classification and Data Analysis