Lectures on Bayesian Machine Learning

@ The Winter Conference in Statistics, Hemavan, Sweden, March 11-14, 2019.

---

Lecturer: Mattias Villani
Positions: Professor of Statistics at Stockholm and Linköping University.
Research: Computationally efficient Bayesian methods for inference, prediction and decision making with flexible probabilistic models.
Teaching: Bayesian Learning, Introduction to Machine Learning, Advanced Machine Learning, Text Mining, Machine Learning for Industry, Probability and Statistics for Machine Learning etc
Current Application Areas: Transportation, Neuroimaging, Robotics, Econometrics and Software Development.
Web: Web page
Google Scholar: Profile
Genealogy: Mathematics Genealogy Project

---

Lecture 1 - The Bayesics

Reading: Slides | Chapter 1 in Pattern Recognition and Machine Learning.
Code: Bayesian regression
Software: RStan

---

Lecture 2 - Gaussian Process Regression

Reading: Slides | Chapter 2 and 4-5 in Gaussian Processes for Machine Learning.
Code: GP for LIDAR data | demo hyperparameter optimization Python
Software: GP fit package R | GPStuff for Matlab | GPy for Python

---

Lecture 3 - Making Use of Gaussian Processes

Reading: Slides | Chapter 3 in Gaussian Processes for Machine Learning | BayesOpt Paper.
Software: rBayesianOptimization package R | GPyOpt | bayesopt Matlab

---

Lecture 4 - Topic Models for Text

Reading: Slides | original topic model article | Topic model intro | Intro PhD thesis Måns Magnusson
Software: R package topic models

---