Active Causal Learning for Decoding Chemical Complexities with Targeted Interventions
This work addresses a problem in chemistry and materials science by improving design tasks with targeted interventions, though it is incremental as it builds on existing active learning methods.
The paper tackles the challenge of predicting molecular properties across datasets by introducing an active learning approach that identifies causal relationships through strategic sampling, enabling optimization in unseen chemical spaces, as demonstrated on the QM9 dataset for finding molecules with large dipole moments.
Predicting and enhancing inherent properties based on molecular structures is paramount to design tasks in medicine, materials science, and environmental management. Most of the current machine learning and deep learning approaches have become standard for predictions, but they face challenges when applied across different datasets due to reliance on correlations between molecular representation and target properties. These approaches typically depend on large datasets to capture the diversity within the chemical space, facilitating a more accurate approximation, interpolation, or extrapolation of the chemical behavior of molecules. In our research, we introduce an active learning approach that discerns underlying cause-effect relationships through strategic sampling with the use of a graph loss function. This method identifies the smallest subset of the dataset capable of encoding the most information representative of a much larger chemical space. The identified causal relations are then leveraged to conduct systematic interventions, optimizing the design task within a chemical space that the models have not encountered previously. While our implementation focused on the QM9 quantum-chemical dataset for a specific design task-finding molecules with a large dipole moment-our active causal learning approach, driven by intelligent sampling and interventions, holds potential for broader applications in molecular, materials design and discovery.