Multitask Learning with No Regret: from Improved Confidence Bounds to Active Learning

TL;DR

This paper develops new confidence intervals for multitask learning in the agnostic setting, enabling improved online regret bounds and active learning strategies that adapt to task similarity without prior knowledge.

Contribution

It introduces novel multitask confidence intervals, an adaptive online learning algorithm, and a multitask active learning setup with theoretical guarantees and empirical validation.

Findings

Improved regret bounds depending on task similarity

Adaptive algorithm that does not require prior knowledge of task similarity

Validated effectiveness on synthetic and drug discovery data

Abstract

Multitask learning is a powerful framework that enables one to simultaneously learn multiple related tasks by sharing information between them. Quantifying uncertainty in the estimated tasks is of pivotal importance for many downstream applications, such as online or active learning. In this work, we provide novel multitask confidence intervals in the challenging agnostic setting, i.e., when neither the similarity between tasks nor the tasks' features are available to the learner. The obtained intervals do not require i.i.d. data and can be directly applied to bound the regret in online learning. Through a refined analysis of the multitask information gain, we obtain new regret guarantees that, depending on a task similarity parameter, can significantly improve over treating tasks independently. We further propose a novel online learning algorithm that achieves such improved regret without knowing this parameter in advance, i.e., automatically adapting to task similarity. As a second key application of our results, we introduce a novel multitask active learning setup where several tasks must be simultaneously optimized, but only one of them can be queried for feedback by the learner at each round. For this problem, we design a no-regret algorithm that uses our confidence intervals to decide which task should be queried. Finally, we empirically validate our bounds and algorithms on synthetic and real-world (drug discovery) data.