Diego A. Mesa – posts

All publications are available on my Google Scholar. Instead, this page gives a working draft of an interactive Machine Learning with Python+Numpy textbook I am working on.

A Bottom-Up Introduction to Neural Networks

From 1-D Forward propagation to vectorized Backprop and everything in between

neural networks

linear algebra

numpy

In this notebook we will walk through the forward and backward computation directions (fowardprop/backprop) of neural networks.

An Introduction to Optimization

From Gradients, to Stochastic Gradient Descent with Momentum and ADAM and everything inbetween!

optimization

numpy

In this notebook we will walk through the several gradient-based optimization techniques from scratch.

An Introduction to Markov Chains, MCMC and The Gibbs Sampler

From Probability to Markov Chain Monte Carlo and the Gibbs Sampler!

probability

mcmc

In this notebook, we begin by reviewing some of the basics of probability and Bayes rule, then make out way into markov chains, and monte-carlo sampling…