Topic outline

  • General

    Description - Machine learning for materials science:

    Machine learning (ML) techniques enable us to infer relationships from a large amount of seemingly uncorrelated input data. Their predictive power has made them central to product development in IT and we already use them in daily life (Amazon, Netflix, etc.). Physical sciences have been slow to capitalize on the promise of ML, even though their computational implementation is suited to modern simulation techniques. Materials science has recently benefited from a number of ML applications to materials discovery and design (featuring neural networks, genetic algorithms, regression methods, compressed sensing and Bayesian optimisation), that promise to accelerate development of novel technologies. Machine learning for materials science is an exciting new discipline that is now being taught at Aalto University.

    "Machine learning in materials science" is a project-led lecture course for graduate students who wish to acquire key skills in this cross-disciplinary research field. Introductory lectures on materials science and machine learning will be followed by tutorial exercises, student-led seminars and a light research project. The research work will be carried out in mixed teams (free choice of topic) and provides ideal opportunities for learning on realistic materials science datasets (experimental or computational). The target of the project work is to get hands-on experience in this field and explore the performance of different ML methods on different datasets, which already constitutes an important contribution to the field and the career of course participants.

    Course level

    The course is intended for students who have completed their Bachelor's degree and have a basic understanding of machine learning or material science and a keen interest interdisciplinary science. 

    Credits

    3 or 5 ECT are awarded for the course. 

    Assessment

    The course grade is pass/fail. It is possible to take the course with two levels of difficulty:

    • level 1: attending classes and tutorials, surveying the field and giving a seminar (pass with 3 ECT). Attendance: classes, tutorials, seminars: can miss 2 out of 9 sessions.
    • level 2: level 1, with seminar work replaced by participating in the project work (pass with 5 ECT). Attendance: classes, tutorials, seminars, project presentations: can miss 2 out of 12 sessions.
    Course structure and workload 

    Period 1 and 2

    • 5 x 2 h introductory lectures on machine learning in materials science
    • 2 x 4h hands-on tutorial sessions
    • 2 x 2h student seminar presentations
    • project work in small teams
    • 3 x 2h project checkpoint contact sessions
    There is no homework for the course and no final exam.

    Teachers

    Learning outcomes

    After completion of the course you will be able to:

      • Identify research questions in material science (MS) that can be solved by machine learning (ML)
      • Understand different types of MS datasets for ML
      • Perform basic data analysis of datasets
      • Select a suitable MS data representation as input for ML
      • Consider which ML methods might be best for tackling different MS problems
      • Assess and improve the performance of the ML model
      • Carry out a computational project on ML for MS
      • Critically comment on ML applications in MS (on quality of data analysis, suitability of chosen ML method, quality of assessment of ML performance, etc).


    Course dates

    8.9-4.12.2020