Fair Resource Allocation in Multi-Task Learning

TL;DR

This paper introduces FairGrad, a novel optimization method for multi-task learning that addresses conflicting gradients by applying fairness principles, leading to improved performance across various benchmarks.

Contribution

The paper proposes FairGrad, a new optimization approach that incorporates fairness into multi-task learning, with theoretical guarantees and extensive empirical validation.

Findings

FairGrad achieves state-of-the-art results on multiple benchmarks.

Incorporating $$-fairness improves performance of existing MTL methods.

The method provides a flexible framework for emphasizing specific tasks.

Abstract

By jointly learning multiple tasks, multi-task learning (MTL) can leverage the shared knowledge across tasks, resulting in improved data efficiency and generalization performance. However, a major challenge in MTL lies in the presence of conflicting gradients, which can hinder the fair optimization of some tasks and subsequently impede MTL's ability to achieve better overall performance. Inspired by fair resource allocation in communication networks, we formulate the optimization of MTL as a utility maximization problem, where the loss decreases across tasks are maximized under different fairness measurements. To solve this problem, we propose FairGrad, a novel MTL optimization method. FairGrad not only enables flexible emphasis on certain tasks but also achieves a theoretical convergence guarantee. Extensive experiments demonstrate that our method can achieve state-of-the-art performance among gradient manipulation methods on a suite of multi-task benchmarks in supervised learning and reinforcement learning. Furthermore, we incorporate the idea of $\alpha$-fairness into loss functions of various MTL methods. Extensive empirical studies demonstrate that their performance can be significantly enhanced. Code is provided at \url{https://github.com/OptMN-Lab/fairgrad}.