Brookhaven AIMS Tutorial: Dimensionality Reduction

Tuesday 30 May 2023, 12:00 → 13:00 US/Eastern

In 2023, we have expanded the Brookhaven AIMS series to include also hands-on tutorials organized by the BNL Computational Science Initiative, which are aimed at researchers, educators and students new to machine learning and artificial intelligence. The tutorial presentations are designed for a general audience, and minimal prior experience will be required.

 

12:00 → 13:00 Dimensionality Reduction

Abstract: Real world data can often be high dimensional, meaning having many features. From images with thousands of pixels, to scientific data with many properties, it is often productive to reduce the "dimensionality" of your datasets in statistically meaningful ways, either before applying predictive methodologies, or simply for helping with visualization and analysis. In particular, we will focus on Principal Component Analysis, which is arguably the simplest and most well known of these methods. In this tutorial, you will learn the basics of dimensionality reduction, the theory behind it and how to apply it to a variety of scientific problems.
GitHub link: https://github.com/matthewcarbone/AIML-tutorials

Speaker: Matthew Carbone (Brookhaven National Laboratory)

dimensionality_reduction.mp4

GitHub Repository