Day 3
The following summarizes the content coverage and progression for the Day 3 Machine Learning workshop at ACS MARM 2022.
Overview & Environment Setup (10 min.)
Roadmap
Objectives for workshop
Outline of planned activities
Instructions for Notebooks
Uncertainty Quantification and Probabilistic Modeling (25 min.)
Roadmap
Gaussian Process Regression
Interactive building simple GPR
Uncertainty with Neural Networks
General Discussion
Bayesian optimization, Active Learning, Design (20 min.)
Roadmap
General framework
Specific considerations
Topical examples from literature
Break (15 min.)
Roadmap
Drink coffee
Make friends
Complete exercises
Activity 2: Active Learning Team Challenge (45 min.)
Roadmap
Discussion of exercise
Brief overview of leruli
Conceptualization with linear regression
Exercise on data viewing/inspection
Excercise on predicting solubility with linear model
Break (15 min.)
Roadmap
Drink coffee
Make friends
Complete exercises
Explainable AI (30 min.)
Roadmap
General concepts: what is an explanation?
Feature importance
Shapley Values
Counterfactuals and
exmolSurrogate models
Activity 3: Explaining Black-box models
Roadmap
Discussion on exercise goals/objectives
Exercise time