Latest

This post provides theoretical foundation behind the Weingarten map for curvature estimation along with its Python implementation for 2D images.

Explaning key building blocks and main mathematical ideas behind the variational Auto Encoders

Lifting of raw data to higher dimensional feature space while maintaining localily.

Introduction into expectation maximization algorithm for fitting a mixture of parametric probabalistic models to data

This block covers ideas on energy based models starting physics via Gibbs distribution abd Restricted Boltzmann Machine.