Tackling Multiple Tasks with One Single Learning Framework

TL;DR

This paper introduces HTAN, a flexible framework for deep multi-task learning that dynamically models task relationships over time, improving performance in sequential learning scenarios.

Contribution

The paper proposes HTAN, a novel hierarchical and temporal framework with adaptive activation functions and regularization techniques for better knowledge sharing in sequential DMTL.

Findings

HTAN outperforms state-of-the-art methods in sequential DMTL tasks.

The framework effectively models time-variant task relationships.

Regularization enhances the learning of task hierarchies.

Abstract

Deep Multi-Task Learning (DMTL) has been widely studied in the machine learning community and applied to a broad range of real-world applications. Searching for the optimal knowledge sharing in DMTL is more challenging for sequential learning problems, as the task relationship will change in the temporal dimension. In this paper, we propose a flexible and efficient framework called HierarchicalTemporal Activation Network (HTAN) to simultaneously explore the optimal sharing of the neural network hierarchy (hierarchical axis) and the time-variant task relationship (temporal axis). HTAN learns a set of time-variant activation functions to encode the task relation. A functional regularization implemented by a modulated SPDNet and adversarial learning is further proposed to enhance the DMTL performance. Comprehensive experiments on several challenging applications demonstrate that our HTAN-SPD framework outperforms SOTA methods significantly in sequential DMTL.