Wei, T., Liu, J., Gupta, A., Tan, P. S., & Ong, Y.-S. (2024). Bayesian Forward-Inverse Transfer for Multiobjective Optimization. In Parallel Problem Solving from Nature – PPSN XVIII (pp. 135–152). Springer Nature Switzerland. https://doi.org/10.1007/978-3-031-70085-9_9
Abstract:
We present an evolutionary optimizer incorporating knowledge transfer through forward and inverse surrogate models for solving multiobjective problems, within a stringent computational budget. Forward knowledge transfer is employed to fully exploit solution-evaluation datasets from related tasks by building Bayesian forward multitask surrogate models that map points from decision to objective space. Inverse knowledge transfer via Bayesian inverse multitask models makes possible the creation of high-quality solution populations in decision space by mapping back from preferred points in objective space. In contrast to prior work, the proposed method can improve the overall convergence performance to multiple Pareto sets by fully exploiting information available for diverse multiobjective problems. Empirical studies conducted on benchmark and real-world multitask multiobjective optimization problems demonstrate the faster convergence rate and enhanced inverse modeling accuracy of our algorithm compared to state-of-the-art algorithms.
License type:
Publisher Copyright
Funding Info:
This research / project is supported by the National Research Foundation, Singapore and DSO National Laboratories - AI Singapore Programme - ”Design Beyond What You Know”: Material-Informed Differential Generative AI (MIDGAI) for Light-Weight High-Entropy Alloys and Multi-functional Composites (Stage 1a)
Grant Reference no. : AISG2-GC-2023-010
This research / project is supported by the Agency for Science, Technology and Research - Singapore RIE2025 Manufacturing, Trade and Connectivity (MTC) Industry Alignment Fund-Pre-Positioning - Distributed Smart Value Chain programme
Grant Reference no. : M23L4a0001
Description:
This is a post-peer-review, pre-copyedit version of an article published in Lecture Notes in Computer Science. The final authenticated version is available online at: http://dx.doi.org/10.1007/978-3-031-70085-9_9