This repository contains the implementations of the lab assignments done for the MSc Course Multivariate Data Analysis at TU Delft. All the algorithms are implemented from scratch.
The first lab assignment contains the implementation of the simple online linear regression algorithm where mean and covaraince of the posterior are updated with every new data sample..
The second lab assignment contains the implementation of different algorithms for computing the posterior when closed-form solution isn't possible: Laplace approximation, Markov Chain Monte Carlo, Gibb's Sampling.
The third lab assignment contais some example realisations of Gaussian Processes with different kernels and the algorithm for Gaussian Process classification using Gibb's Sampling.