Machine Learning for Materials and Chemistry

Aims to be a elective Master module, can be followed by Bachelor students. Developed at University of Kassel.

Slides

• Meta
• Definition and Overview
• Problem Classes
• Learning Workflow
• Representations
• Decision Trees
• k-Nearest Neighbor
• Kernel Ridge Regression
• Assessing Models
• Cross-Validation
• Neural networks
• Language Models

Exercises

Meant to both deepen understanding of the lecture content, but also to boost Python skills.

Week 1: Linear regression and model stability
Week 2: Simple representations
Week 3: Loss functions
Week 4: Cutoffs and expansions
Week 5: Decision trees and sklearn
Week 6: Kernels
Week 7: Hyperparameters
Week 8: Learning vs performance
Week 9: Cross-Validation
Week 10: Neural networks