Computational Materials Design

What it is about

In this project-oriented master's module you will learn how high-throughput simulations and data-driven methods combine to predict material properties. By the end you will be able to implement ML models in Python, analyze datasets from the materials sciences, and assess the quality of different models.

What you will be able to do

Develop Python programs
Perform basic data analyses
Understand types of datasets in the materials sciences
Identify scientific questions for ML
Assess the quality of ML models
Compare different ML methods
Understand the importance of data-driven methods for material properties

What the module consists of

VorlesungLecture of content by lecturers and subsequent discussion; visualization using presentations
Übungpractical application of Python for ML and use of open-source libraries in exercises and project work
Projektarbeitenpractical implementation and comparison of various ML applications in Python

Teaching method

VorlesungenConveying subject-specific fundamentals and discussion of central concepts
Übungenpractical experience with Python, scikit-learn and PyTorch to implement ML solutions
ProjektarbeitApplication and deepening of the methods on real tasks

What you should already know

Grundlagen der Programmierungrecommended
Grundkenntnisse der Festkörperphysik, Mathematik, Material- und Ingenieurswissenschaftenrecommended
EI04024: Python for Engineering Data Analysisrecommended
EI71066: Simulation of Semiconductor propertiesrecommended

Assessment

Project workHausaufgabePresentation

The module assessment consists of two equally weighted parts: projects (50%) and assignments (50%). In two accompanying projects you implement ML applications in Python, present the implementation and results in a ~15-minute presentation each, and discuss them with fellow students. Additionally there are two assignments in which theoretical ML and data science concepts are developed and summarized in roughly two-page essays.

Reading list

Y. Cheng, T. Wang, G. Zhang - Artificial Intelligence for Materials Science
A. Géron - Hands-On Machine Learning with Scikit-Learn, Keras, and TensorFlow

Dates

Lecture with exerciseComputational Materials Design2 groups to choose from

ATue12:00–13:30
03.5901.022, Seminarraum (5901.03.022)
15× · 13.10.–02.02.
- 13.10.
- 20.10.
- 27.10.
- 03.11.
- 10.11.
- 17.11.
- 24.11.
- 01.12.
- 08.12.
- 15.12.
- 22.12.
- 12.01.
- 19.01.
- 26.01.
- 02.02.
BTue13:45–16:00
03.5901.022, Seminarraum (5901.03.022)
15× · 13.10.–02.02.
- 13.10.
- 20.10.
- 27.10.
- 03.11.
- 10.11.
- 17.11.
- 24.11.
- 01.12.
- 08.12.
- 15.12.
- 22.12.
- 12.01.
- 19.01.
- 26.01.
- 02.02.

From the current semester, not binding. You attend one of several groups; the timetable automatically suggests the one with the fewest clashes.

Filters