Computational Bayesian Time Series is a combination of two short courses. In the first part, computational Bayesian topics that are relevant for Deep learning will be introduced. The second half covers Bayesian time series, which will be the base knowledge for advanced topics in reinforcement learning.

In the first half of the course, computational Bayesian discusses how MCMC can help extract the most information from a set of limited sample data. Topics discussed will be following

• Gibbs Sampling
• Metropolis-Hastings
• Variational Inference

All of which will be the key elements of Advanced Deep Learning.

A leading theme, in the following second half of the course, is advanced time series topics that are closely related to MCMC, which are,

• Generalized Method of Moments (GMM) – in time series context
• Vector AutoRegression (VAR)

Above topics will be re-introduced in Advanced Reinforcement Learning