Classifying materials for next generation nanotechnologies with machine learning



Discovering, design and choosing the right material for a chemical or engineering application typically involves developing a structure/property relationship which relates the structural features of the system to a functional property as the target label.  This can be very useful to inform engineers what type of material to make, provided they have exquisite control and can make the perfect material every time.  In the case of nanomaterials (which are larger than molecules, but smaller than materials; only millionth of a millimetre in size) it is not possible to make pure samples and distributions and mixtures are always present.  In this case it is better to predict the performance of Classes of structures, rather than individual structures than may be impossible to replicate. In this project you will use a series of supervised and unsupervised machine leaning methods to predict the classes in a nanomaterials data set, report in the accuracy, precision and recall, and test the generalisability of the model using learning curves.  All programming will be done in python, with packages available in scikit-learn. Data sets will be provided.


Python programming and an interest in data science and machine learning is essential.

Please double check my email when enquiring to ensure your message finds the right professor.


machine learning, classification, data science, materials, nanotechnology, python

Updated:  10 August 2021/Responsible Officer:  Head of School/Page Contact:  CECS Webmaster