Revolutionary Machine Learning Predicts Material Properties from Minimal Data | Bengaluru News

BENGALURU: Researchers at the Indian Institute of Science (IISc) and University College London have developed a machine learning method to predict material properties using limited data. They believe this can aid in the discovery of materials with desired properties, such as semiconductors.
“In recent years, materials engineers have turned to machine learning models to predict which types of materials can possess specific properties such as electronic band gaps, formation energies, and mechanical properties, in order to design new materials,” IISc said.
It added that, however, data on material properties – which is needed to train these models – is limited because testing materials is expensive and time-consuming. This prompted researchers led by Sai Gautam Gopalakrishnan, Assistant Professor at the Department of Materials Engineering, IISc, to work on addressing this challenge.
In a new study, they found an efficient way to use a machine learning approach called transfer learning to predict the values of specific material properties.
In transfer learning, a large model is first pre-trained on a large dataset and then fine-tuned to adapt to a smaller target dataset. “In this method, the model first learns to do a simple task like classifying images into, say, cats and non-cats, and is then trained for a specific task, like classifying images of tissues into those containing tumours and those not containing tumours for cancer diagnosis,” Gopalakrishnan said.
Their transfer learning-based model demonstrated superior performance compared to models trained from scratch, Reshma Devi, the study’s lead author, said. The team’s Multi-property Pre-Training (MPT) framework achieved remarkable results by simultaneously training on seven different bulk 3D material properties. Most notably, the model successfully predicted band gap values for 2D materials it never encountered during training.
Looking ahead, the technology shows promise for battery development by predicting ion movement within electrodes. “This research could contribute significantly to India’s semiconductor manufacturing initiatives by helping predict material defects,” Gopalakrishnan said.