Back in 2022 I really started to get interested in data science. I always wanted to learn more about bayesian statistics, and how to use it in practice beyond the classic expression of the Bayes formula we learn in class. So I dig into a bayesian statistics introduction book, and with the help of some youtube videos (from the excellent channel ritvikmath), I created my first bayesian statistics algorithm.
Creating a Markov-chain Monte-Carlo type algorithm to estimate the posterior of a bayes formula given any prior. I followed the Metropolis-Hsting algorithm principle.
- echantillonage_preferentiel_v1.py : Training script for the accept-reject sampling
- MCMC_v1.py : First iteration of the MCMC algorithm
- MCMC_v2.py : Second iteration of the MCMC algorithm