Back to Main Page

Here I plan to explore concepts and ideas from the world of data analysis and machine learning (with a strong focus on topological and geometric approaches), make my thoughts and my understanding of those concepts accessible to myself in the future, and simultaneously share these things with others (because sharing is good).

The intent is not to target the most broad audience. Usually my focus lies on the core ideas behind the algorithms rather than their most performant implementations.

Table of contents:

The Metropolis-Hastings Algorithm — MCMC methods (1) 19 July 2021
The Chinese Restaurant Process — Derivation from finite Dirichlet mixtures 27 May 2021
Finite Sample Expressivity of Neural Networks 12 January 2017
Compression after Lempel and Ziv 17 May 2016
Sparse Distributed Representions and Witness Complexes 22 February 2016
Mapper: (Naive) Python Implementation 22 February 2015
K-Means clustering and Lloyd's algorithm 23 January 2015
Mapper — A discrete generalization of the Reeb graph 8 December 2014
OPTICS — Density based clustering (2) 23 November 2014
DBSCAN — Density based clustering (1) 17 November 2014