Коллоквиум : On robust mean estimation and k means clustering

Никита Животовский In this talk we consider the robust algorithms for the kmeans clustering problem where a quantizer is constructed based on N independent observations. We start with an overview of the methods of robust statistics. First, we discuss the medianofmeans estimator. This simple estimator allows us to evaluate the mean of some heavytailed distribution as if this distribution was Gaussian. We discuss some extensions to the multivariate case. In the context of clustering, we present the media