Multi-Objective Learning to Predict Pareto Fronts Using Hypervolume Maximization
arXiv:2102.04523
Abstract
Real-world problems are often multi-objective with decision-makers unable to specify a priori which trade-off between the conflicting objectives is preferable. Intuitively, building machine learning solutions in such cases would entail providing multiple predictions that span and uniformly cover the Pareto front of all optimal trade-off solutions. We propose a novel approach for multi-objective training of neural networks to approximate the Pareto front during inference. In our approach, the neural networks are trained multi-objectively using a dynamic loss function, wherein each network's losses (corresponding to multiple objectives) are weighted by their hypervolume maximizing gradients. We discuss and illustrate why training processes to approximate Pareto fronts need to optimize on fronts of individual training samples instead of on only the front of average losses. Experiments on three multi-objective problems show that our approach returns outputs that are well-spread across different trade-offs on the approximated Pareto front without requiring the trade-off vectors to be specified a priori. Further, results of comparisons with the state-of-the-art approaches highlight the added value of our proposed approach, especially in asymmetric Pareto fronts.
arXiv commentT.M.D. and M.G. contributed equally. Changes in new version (v2): improved method, added comparison to method described in old manuscript (v1), added experiments, added appendix, revised text
References in corpus (13)
- An Overview of Multi-Task Learning in Deep Neural Networks
- A large annotated medical image dataset for the development and evaluation of segmentation algorithms
- Multi-Task Learning as Multi-Objective Optimization
- Differentiable Expected Hypervolume Improvement for Parallel Multi-Objective Bayesian Optimization
- Gradient Surgery for Multi-Task Learning
- Pareto Multi-Task Learning
- Learning the Pareto Front with Hypernetworks
- Efficient Continuous Pareto Exploration in Multi-Task Learning
- Multi-objective training of Generative Adversarial Networks with multiple discriminators
- Controllable Pareto Multi-Task Learning
- Random Hypervolume Scalarizations for Provable Multi-Objective Black Box Optimization
- Multitask Learning with Single Gradient Step Update for Task Balancing
- Single-Solution Hypervolume Maximization and its use for Improving Generalization of Neural Networks