(30 min) This post demonstrates how Test Driven Development (TDD) applies to developing Machine Learning frameworks such as Tensorflow


(30 min) Part 2 of 2 series on variational inference. This part dives into the more practical black-box variational inference. We discuss the REINFORCE algorithm and gradient variance reduction techniques (including the neural baseline).


(30 min) Literature review of popular methods in few-shot learning. With a focus on metric-learning and meta-learning.


(30 min) Part 1 of 2 series on variational inference. This part first introduces high-level concepts and the mean-field approximation. We then demonstrate the mean-field theory on a Bayesian Gaussian Mixture Model (Bayesian GMM).


(45min) This post covers the in-depth theory regarding the EM algorithm, with python implementation for a Gaussian Mixture Model from scratch.