Machine Learning theory and methods

Syllabus

The syllabus covers the following topics organized in the following 5 units:

Introduction

Introduction and background to the topic

Fundamentals

1. Machine Learning: concept, categories, applicability, features and extraction
2. Data: data structures, use, storage (FAIR principles), platforms and databases
3. Features: importance and extraction, examples, workflows
4. Explainability: use cases and critical assessment
5. Ensemble learning: boosting, trapping

Methods and algorithms

1. Recap: Unsupervised and supervised learning: selected methods (principal component analysis, kmeans, DBSCAN; regression)
2. Neural Networks: mathematical build-up, hyperparameters, selected networks (long short-term memory, graph neural networks), autoencoders, physics-informed neural networks
3. Generative models: concepts and differences from non-generalized models, examples (variational autoencoders, general adversial networks)
4. Hybrid models: concepts, examples
5. Large language models: transformers, tokenization

Design of soft materials & (bio)molecules

1. Structure: proteins (AlphaFold), molecules, fingerprinting/featurization, selected prediction models
2. Properties: embeddings, use cases

Machine Learning and computer simulations

1. Potentials: development generations, distinct energy descriptions
2. Interactions: short-, medium-, and long-range

Learning Outcomes

Upon successful completion of the course students will be able to

• Gain experience with data and databases and be aware of the fairness in terms of generating, using, and archiving data.
• Become familiar with advanced concepts of Machine Learning with emphasis in Natural Sciences and Engineering.
• Be able to critically assess the applicability, explainability, and generalizability of Machine Learning methods.
• Understand the conceptual differences, strengths & weaknesses of standard with respect to generative learning models.
• Be able to critically compare Machine Learning schemes and algorithms.
• Be able to express a Machine Learning algorithm in form of a pseudocode.
• Express solid-state structures and molecules as embeddings and be able to provide examples and ideas of molecular fingerprints.
• Understand the concept of Machine Learning potentials and explain their applicability.
• Gain a practical hands-on experience with advanced Machine Learning algorithms.

Recommended Bibliography

• I. Goodfellow, Y. Bengio, A. Courville, Deep Learning (Adaptive Computation and Machine Learning series), The MIT Press, USA (2016)
• M.P. Deisenroth, A. A. Faisal, C.S. Ong, Mathematics for Machine Learning, Cambridge University Press (2020) 
• K.P. Murphy, Machine Learning a probabilistic perspective, The MIT Press (2012) 
• S. Shalev-Shwartz, S. Ben-David, Understanding Machine Learning, Cambridge University Press (2014) 
• M. Erdmann, J. Glombitza, G. Kasieczka, U. Klemradt, Deep Learning for Physics Research, World Scientific (2021).