Inverse Design of Nanoparticle Catalysts using Multi-target Machine Learning



Inverse design, where we can prescribe the materials attributes based on a desired target property label, is a singular ambition of data-driven material design, and if successful will finally enable market-pull innovation as opposed to conventional technology-push.  This is extremely challenging however, since there could be numerous combinations of materials characteristics that present the same properties and predicting a traditional “structure/property relationship” does not distinguish between them.  It is even more challenge in nanomaterials design since the design space is even greater, and inverse property/structure relationships will typically need to encompass multi-functionality.  Recently a solution to this problem has been devised based on multi-target machine learning that draws upon this multi-functionality and predicts the exact nanoparticle structure that will deliver a set of desirable properties simultaneously, with a fault tolerance.  The method focusses the outcome on the most important characteristics in an entirely data-driven way, and with comparable accuracy and generalizability as traditional forward structure/property machine learning predictions.  This approach works very well for ordered materials, but its efficacy for highly disordered and complex non-crystalline structures, such as metal nanoparticles using for catalysis, is unknown.  In this project you will apply and test this multi-target inverse design method to data describing metal nanoparticle catalysts and develop a model to inform experimental and manufacturing strategies.   The results can be made actionable using statistical leaning such as Bayesian networks.  The data set will be provided.


To use multi-target regrsssion for inverse design of ordered and disordered nanoparticle electrocatalysts


Python programming and experience in data science and machine learning is essential (such as COMP3720, COMP4660, COMP4670, COMP6670, COMP8420).  Familiarity with platforms such as scikit-learn, Pytorch, Tensorflow and Keras is desirable.


This can be a 12cp or 24cp project.


machine learning, materials informatics

Updated:  10 August 2021/Responsible Officer:  Dean, CECS/Page Contact:  CECS Marketing