The basic idea of Q-learning is that 'the value of a certain state (Q value) is determined by the reward obtained and the value of the state at the next point in time', and is expressed by the ...

This course offers a rigorous yet practical exploration of Bayesian reasoning for data-driven inference and decision-making. Students will gain a deep understanding of probabilistic modeling, and ...

We develop novel methods to make Bayesian inference more efficient, scalable, and practical. This includes work on variational methods, Monte Carlo algorithms, and techniques for handling complex ...