Date Thesis Awarded

4-2020

Document Type

Honors Thesis

Degree Name

Bachelors of Science (BS)

Department

Mathematics

Advisor

Charles Johnson

Committee Members

Gexin Yu

Martin White

Abstract

The Nonnegative Matrix Factorization (NMF) problem has been widely used to analyze high-dimensional nonnegative data and extract important features. In this paper, I review major concepts regarding NMF, some NMF algorithms and related problems including initialization strategies and near separable NMF. Finally I will implement algorithms on generated and real data to compare their performances.

Share

COinS