Quantum Machine Learning for Predictive Chemistry

The project develops a quantum machine learning (QML) approach to efficiently predict bond dissociation enthalpies (BDEs)—a key molecular property essential for understanding chemical reactivity and pollutant degradation, particularly in persistent substances like PFAS. Traditional BDE calculations are computationally intensive, limiting data availability for machine learning. To overcome this, we apply Quantum Gaussian Process Regression (QGPR), a hybrid quantum-classical method that combines quantum computing and Bayesian inference to make accurate predictions with limited data. By integrating QGPR into an active learning framework, we aim to reduce computational costs and accelerate the identification of chemically informative molecules. This work highlights the potential of QML to advance predictive chemistry and environmental science.