Some topics of Bayesian Inference

by Alberto Ramos, Bryan Zaldivar

Monday 4 Apr 2022, 11:00 → Friday 8 Apr 2022, 13:00 Europe/Madrid

The course will consist of 10 hours, 2h per day, from 11:00 to 13:00. The outline is the following:

1. Fundamentals of Bayesian Inference. Laplace Approximation 

2. Variational Inference; Kullback-Leibler divergence. Working example with Tensorflow v2 (python)

3. Inference in function space: Gaussian Processes, Implicit Processes.

4. Markov Processes; Markov Chain Monte Carlo; Hybrid (a.k.a. Hamilton) Monte Carlo. 

5. Approximate Bayesian Computation (ABC). Likelihood-free method. Normalizing flows.

        

Many of the topics will be accompanied by coding examples.

 

****** DAY, TIME, ROOM ***

- Monday, 11:00 - 13:00.  1001-Primera-135 - Nave Exp. Sala de Audiovisuales

- Tuesday, 9:30 - 11:30.  1001-Primera-1-1-1 - Paterna. Seminario

- Wednesday, 11:00 - 13:00. 1001-Primera-1-1-1 - Paterna. Seminario

- Thursday, 11:00 - 13:00. 1001-Primera-1-1-1 - Paterna. Seminario

- Friday, 11:00 - 13:00. 1001-Primera-1-1-1 - Paterna. Seminarioles

lecture1_Bayesian_sequential_learning.ipynb lecture2_differentiation_with_tensorflow.ipynb lecture2_VI_bayesian_poly_tensorflow.ipynb lecture3_GP_regression.ipynb lecture4_mcmc_exercises.pdf lecture4_mcmc_hmc.pdf lecture4_mcmc_pi.pdf lecture5_likelihood-free_polynomial.ipynb project_period.txt project_statement.pdf project_trajectory.txt