This project demonstrates Stochastic Variational Inference in the TrueSkill Model, using Bayesian inference in large models with continuous latent variables and Markov Chain Monte Carlo on tennis data sets.

Implementations include:

• Project.toml packages for the Julia environment.
• autograd_starter.py Python with autograd.
• tennis_data.mat dataset containing outcomes of tennis games.
• src.jl source code in Julia
• Julia-True-Skill-Model-Using-Markov-Chain-Monte-Carlo-Final.ipynb the jupyter notebook

Note: this project is part of of the assignment from Statistical Methods for Machine Learning II at the Univeristy of Toronto.