Bayesian

Skeleton notes for conditional probability and Bayes' theorem

EM method, expectation-maximization algorithm, is an inspiring iterative method to find the log-likelihood by introducing some intermediate variable such as responsibility.

Topics Variational Inference Normalizing Flow Variational Inference with Normalizing Flows

Topic: uncertainty in deep learning References: Gawlikowski, J. et al. A Survey of Uncertainty in …

Our topic for this session is Cerqueira V, Torgo L, Mozetic I. Evaluating time series forecasting …

Our topic for this session is Inferring causal impact using Bayesian structural time-series models …