Machine Learning for Materials Research: Bootcamps and Workshops

Summary

Over the last several years machine learning has gained increased attention in the material science community. The ability to perform high throughput in silico experiments to simulate novel materials and their properties lends itself perfectly to the use of machine learning techniques to assist in the discovery of materials with certain desirable properties. The purpose of the bootcamps is to educate members of industry, academia, and national labs on how to use machine learning in their applications.

Description

The 2016 bootcamp consisted of three days of lectures covering data processing, supervised learning and unsupervised learning as well as hands-on exercises using MATLAB covering a range of data analysis topics touching on each of the lecture . Example topics include:

• Identifying important features in complex/high dimensional data
• Visualizing high dimensional data to facilitate user analysis.
• Identifying the fabrication 'descriptors' that best predict variance in functional properties.
• Quantifying similarities between materials using complex/high dimensional data

The hands-on exercises focused on demonstrating practical use of machine learning tools on real materials data. Attendees will learn to analyze a range of data types from scalar properties such as material hardness to high dimensional spectra and micrographs.

URL: https://www.nanocenter.umd.edu/events/mlmr/

NIST Organizers:
Gilad Kusne: https://www.nist.gov/people/aaron-gilad-kusne 
Daniel Samarov: https://www.nist.gov/people/daniel-victor-samarov 

Major Accomplishments

Education

Additional Technical Details

A workshop/bootcamp was given in the Summery of 2016, another is planned for the upcoming Summer of 2017.