Bayesian Estimation: Unsupervised Learning & MCMC

This lecture covers Bayesian estimation for unsupervised learning, focusing on Monte Carlo Markov chains and their relation with Statistical physics. The instructor explains the concept through an example of a Spin Glass Card game, illustrating the goal of splitting a room into two groups based on observed data. Topics include Maximum Likelihood Estimator, Bayes' theorem, statistical inference, Boltzmann measure, and partition function. The lecture also delves into interaction and phase transition in the context of statistical physics.