Voronoi-grid-based Pareto Front Learning and Its Application to Collaborative Federated Learning

Voronoi-grid-based Pareto Front Learning and Its Application to Collaborative Federated Learning

Forty-second International Conference on Machine Learning

10.48550/arXiv.2505.20648

https://openreview.net/forum?id=hrBfufwMzg

Mengmeng Chen, Xiaohu Wu, Qiqi Liu, Tiantian He, Yew-Soon Ong, Yaochu Jin, Qicheng Lao, Han Yu

federated learning, multi-objective optimization, multi-objective evolutionary algorithm

13 July 2025

Chen, M., Wu, X., LIU, Q., He, T., Ong, Y. S., Jin, Y., ... & Yu, H. Voronoi-grid-based Pareto Front Learning and Its Application to Collaborative Federated Learning. In Forty-second International Conference on Machine Learning.

Multi-objective optimization (MOO) exists extensively in machine learning, and aims to find a set of Pareto-optimal solutions, called the Pareto front, e.g., it is fundamental for multiple avenues of research in federated learning (FL). Pareto-Front Learning (PFL) is a powerful method implemented using Hypernetworks (PHNs) to approximate the Pareto front. This method enables the acquisition of a mapping function from a given preference vector to the solutions on the Pareto front. However, most existing PFL approaches still face two challenges: (a) sampling rays in high-dimensional spaces; (b) failing to cover the entire Pareto Front which has a convex shape. Here, we introduce a novel PFL framework, called as PHN-HVVS, which decomposes the design space into Voronoi grids and deploys a genetic algorithm (GA) for Voronoi grid partitioning within high-dimensional space. We put forward a new loss function, which effectively contributes to more extensive coverage of the resultant Pareto front and maximizes the HV Indicator. Experimental results on multiple MOO machine learning tasks demonstrate that PHN-HVVS outperforms the baselines significantly in generating Pareto front. Also, we illustrate that PHN-HVVS advances the methodologies of several recent problems in the FL field.

Attribution-NonCommercial 4.0 International (CC BY-NC 4.0)

This work is supported in part by the National Key R&D Program of China (Grant No. 2024YFE0200500); the National Natural Science Foundation of China (Grant Nos. 62327801 and 62302147); the International Collaboration Fund for Creative Research of the National Science Foundation of China (NSFC ICFCRT) (Grant No. W2441019); the Ministry of Education, Singapore, under its Academic Research Fund Tier 1 (RG101/24); the National Research Foundation, Singapore, and DSO National Laboratories under the AI Singapore Programme (AISG Award No. AISG2RP-2020-019); and the MTI, Singapore, under its AI Centre of Excellence for Manufacturing (AIMfg) (Award W25MCMF014).

https://oar.a-star.edu.sg/communities-collections/articles/22647

https://doi.org/10.48550/arXiv.2505.20648

Institute of High Performance Computing