Abstract
Graph representation of molecular data enables extracting stereoscopic features, with graph neural networks (GNNs) excelling in molecular property prediction. However, selecting optimal hyper-parameters for GNN construction is challenging due to the vast search space and high computational costs. To tackle this, we introduce a hierarchical evaluation strategy integrated with a genetic algorithm (HESGA). HESGA combines full and fast evaluations of GNNs. Full evaluation involves training a GNN with preset epochs, using root mean square error (RMSE) to measure hyperparameter quality. Fast evaluation interrupts training early, using the difference in RMSE values as a score for GNN potential. HESGA integrates these evaluations, with fast evaluation guiding candidate selection for full evaluation, maintaining elite individuals. Applying HESGA to optimise deep GNNs for molecular property prediction, experimental results on three datasets demonstrate its superiority over traditional Bayesian optimisation, Tree-structured Parzen Estimator, and CMA-ES. HESGA efficiently navigates the complex GNN hyperparameter space, offering a promising approach for molecular property prediction.
Original language | English |
---|---|
Title of host publication | Proceedings of the Genetic and Evolutionary Computation Conference 2024 |
Publisher | Association for Computing Machinery |
Pages | 1417-1425 |
Number of pages | 9 |
ISBN (Electronic) | 9798400704949 |
DOIs | |
Publication status | Published - 14 Jul 2024 |
Event | Genetic and Evolutionary Computation Conference 2024 - Melbourne, Australia Duration: 14 Jul 2024 → 18 Jul 2024 |
Conference
Conference | Genetic and Evolutionary Computation Conference 2024 |
---|---|
Abbreviated title | GECCO 2024 |
Country/Territory | Australia |
City | Melbourne |
Period | 14/07/24 → 18/07/24 |
Keywords
- graph neural networks
- hierarchical evaluation strategy
- hyperparameter optimisation
- molecular property prediction
ASJC Scopus subject areas
- Software
- Control and Optimization
- Artificial Intelligence
- Logic
- Computer Science Applications
- Computational Theory and Mathematics