MetaBalance: Improving Multi-Task Recommendations via Adapting Gradient Magnitudes of Auxiliary Tasks

TL;DR

MetaBalance is a novel method that adaptively balances the influence of auxiliary tasks in multi-task recommendation models by adjusting gradient magnitudes, leading to improved recommendation accuracy.

Contribution

It introduces a gradient manipulation technique to dynamically balance auxiliary and target tasks in multi-task learning for recommendations.

Findings

Achieves 8.34% improvement in NDCG@10 over baseline.

Effectively balances auxiliary task influence during training.

Demonstrates robustness across real-world datasets.

Abstract

In many personalized recommendation scenarios, the generalization ability of a target task can be improved via learning with additional auxiliary tasks alongside this target task on a multi-task network. However, this method often suffers from a serious optimization imbalance problem. On the one hand, one or more auxiliary tasks might have a larger influence than the target task and even dominate the network weights, resulting in worse recommendation accuracy for the target task. On the other hand, the influence of one or more auxiliary tasks might be too weak to assist the target task. More challenging is that this imbalance dynamically changes throughout the training process and varies across the parts of the same network. We propose a new method: MetaBalance to balance auxiliary losses via directly manipulating their gradients w.r.t the shared parameters in the multi-task network. Specifically, in each training iteration and adaptively for each part of the network, the gradient of an auxiliary loss is carefully reduced or enlarged to have a closer magnitude to the gradient of the target loss, preventing auxiliary tasks from being so strong that dominate the target task or too weak to help the target task. Moreover, the proximity between the gradient magnitudes can be flexibly adjusted to adapt MetaBalance to different scenarios. The experiments show that our proposed method achieves a significant improvement of 8.34% in terms of NDCG@10 upon the strongest baseline on two real-world datasets. The code of our approach can be found at here: https://github.com/facebookresearch/MetaBalance