BYU

Abstract by Jared Neilson

Personal Infomation


Presenter's Name

Jared Neilson

Co-Presenters

Neal Munson
Jacob Stern

Degree Level

Undergraduate

Co-Authors

Jacob Stern

Abstract Infomation


Department

Computer Science

Faculty Advisor

David Wingate

Title

Semi-Supervised Learning with UMAP Embeddings

Abstract

In machine learning, obtaining sufficient labeled data can be expensive or infeasible. Semi-supervised learning (SSL) tackles this problem by using the distribution of unlabeled data to augment limited labeled data. We apply the unsupervised UMAP algorithm to create fast, descriptive embeddings for data preprocessing. This approach yields improved test accuracy and faster training time on state-of-the-art SSL algorithms. We demonstrate improvements across several synthetic and realistic datasets.