This repository contains jupyter notebooks in response to the assignments given in MMDS course @TU München
- Assignment 1: Duplicate detection in million songs dataset using locality sensitive hashing (LSH).
- Assignment 2: Song recommendation through matrix factorization with 2 optimization implementations. SGD and alternating computation of closed form solution.
- Assignment 3: Song searching with topic specific PageRank algorithm.
- Assignment 4: Spectral clustering on million songs dataset.