Artificial Intelligence for Materials Science

Aim

The purpose of the course is to introduce methods in machine learning and artificial intelligence, and to provide students with the knowledge and skills required to begin applying such methods to solve problems in materials science. Upon completion of the course, students shall be able to:

• implement and use classical machine learning methods as well as neural networks
• select appropriate techniques and algorithms based on given problems
• use materials data from open databases for machine learning
• apply methods in artificial intelligence in a materials science project and present the results.

Content

The course content is structured into six modules.

• Introduction to Python This short module is optional and primarily for students with no prior Python experience. The module offers the essentials of Python required to follow the rest of the course.
• Classical machine learning. This module introduces fundamental machine learning algorithms for regression, classification, clustering, and dimensionality reduction. Alongside algorithmic discussions, we will explore data handling, model training, cost functions, regularization, and model evaluation.
• Neural networks. This module focuses on the basics of neural networks, including components, activation functions, training processes, network design, and regularization techniques. Common network architectures such as feedforward networks, convolutional networks, and autoencoders will be explored.
• Advanced topics This module introduces a few more advanced topics, such as transformers and graph neural networks.For modules 2, 3, and 4, the underlying theory and methodologies are field-independent and broadly applicable. However, as the course is developed primarily with materials science and chemistry in mind, the example and benchmark datasets are drawn largely from these domains.
• Material informatics This module gives an introduction to materials informatics, including how to represent and featurize molecules and materials. It also gives an introduction in how to access materials databases.
• Project work At the end of the course, the students apply AI methods in a materials science project. The students choose a project topic in consultation with the course instructor. The project is presented in a written report and orally.