Physics Informed Machine Learning
Since 2019, the parameters of large deep learning models have increased by over 300 times every 18 months. However, the future ML progress cannot continue simply based on using more data or creating larger models, because the growing gap between the model demand and resource supply is not sustainable. In addition, for many problems in practice, for e.g., predicting protein-drug interactions, catalyst discovery, and advanced material design, we don’t have huge amount of training data to build big models. Ongoing research projects integrate scientific knowledge, physical laws, and simulations with machine learning methods to predict the blueprint or product design of drugs, catalysts, and materials.