<ul><li>Voronoi-grid-based Pareto Front Learning (PHN-HVVS) is introduced for multi-objective optimization in machine learning and federated learning.</li><li>PHN-HVVS utilizes Hypernetworks (PHNs) and Voronoi grids partitioning with genetic algorithm (GA) to improve Pareto front coverage in high-dimensional spaces.</li><li>Experimental results show that PHN-HVVS outperforms existing methods in generating Pareto fronts, benefiting various federated learning tasks.</li><li>Code for PHN-HVVS is available at https://github.com/buptcmm/phnhvvs.</li></ul>

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

Discover more