Mattias Villani

Natural Born Bayesian

Introduction

I am Professor of Statistics and head of the Division of Statistics and Machine Learning (STIMA) at the Department of Computer and Information Science (IDA), Linköping University. I also lead the Machine Learning Research Group at IDA.

My research focuses on developing computationally efficient Bayesian methods for inferenceprediction and decision making using flexible probabilistic models. See my Research page for a list of publications.

Here are some data sets that are currently keeping me and my group busy:

  • tall data – longitudinal data on a quarter of a million Swedish firms
  • wide data – classifying documents and text using thousands of covariates/features
  • massively multivariate data – 4D (space + time) images of brain activity in 100,000 locations in the brain
  • unstructured data – 8 million text documents
  • real-time data – streaming sensor data in robotics.

I currently teach the courses Bayesian learning, Advanced Bayesian Learning (taught irregularly), Advanced Machine Learning, and Probability and Statistics. I also teach the statistical part of the Text mining course.

I try to make notes (i.e. keep a diary for my own fading memory) of recent events on my news page.

    LinkedIn  GitHub - Social coding  Mendeley  Google Scholar

Disclaimer: Any views expressed here are my own and do not necessarily reflect the views of Linköping University, or anyone else for that matter.