Toolbox allows to test and compare methods for Image Completion and Data Completion problems in Matlab. Presented methods use various Nonnegative Matrix Factorization and Tensor decomposition algorithms. It was based on research performed during realization of PhD.

A gentle introduction to nonnegative matrix factorization (NMF), with an application to image compression 🖼️ 🎭 🎨

MATLAB code for stochastic variance reduced multiplicative updates (SVRMU) for NMF 1.0.0

SNMF: Integrated Learning of Mutational Signatures and Prediction of DNA Repair Deficiencies by Goossens S, Tepeli YI, Gonçalves JP

Nonnegative matrix factorization

Nonnegative matrix factorizations using CUDA

Implementation of Bayesian Inference for Nonnegative Matrix Factorisation Models.Computational Intelligence and Neuroscience, 2009

Theme Supervised Nonnegative Matrix Factorization

Minimax NMF

This project shows that companies often give the wrong name to different IT jobs. As a consequence, companies may fail to attract good candidates because an applicant has a significant probability to apply for the wrong job.

My Programming Assignments from the MA 544 Numerical Linear Algebra for Big Data Course from Stevens Institute of Technology

Filtering drums out of music using matrix factorization and CNN ML Model

Nonnegative matrix factorization and matrix allocation algorithm implementations in C++

This project showcases an end-to-end workflow for topic modeling and text analysis using a variety of machine learning and natural language processing techniques. The goal of this project is to extract meaningful topics from a collection of text documents, enabling insights, categorization, and understanding of the underlying themes in the data.

This is an implementation of the following paper (Algorithm1) in Python for image reconstruction. "N. Gillis and S.A. Vavasis, Fast and Robust Recursive Algorithms for Separable Nonnegative Matrix Factorization"

A promising Spotify Wrapped feature using Principal Component Analysis

Selected machine learning projects from courses I have taken during my PhD

Source code for the paper titled "Towards Weak Signal Analysis in Hyperspectral Data: A Semi-supervised Unmixing Perspective"

This notebook analyzes Spotify song data and builds a recommendation system to suggest songs based on user listening behavior.

These codes are written as a part of ECE219 Large Scale Data Mining course at UCLA.