Professor Yanyan He invites you to attend the Master's Project Defense of Lavinia Pieptea

WHEN: Friday, April 1, 2022 at 10:00am via Zoom

"Various Machine Learning (ML) approaches to identify feature importance and their application to material science"

Abstract:

Feature selection is desired to reduce the number of input variables so that the computational cost of modeling can be reduced and, performance of model may be improved. It is well known that machine learning (ML) algorithms can be used for predictions and classifications in many scientific and industrial problems. Some of them can also be used for the purpose of feature selection. In this talk, we will explore various ML approaches (including both linear and nonlinear models) to measure feature importance. Those approaches will be applied to quantitative structure-property relationship analysis of calcium aluminosilicate glasses from the material science domain.