Deep Multi-Task Learning with Shared Memory

TL;DR

This paper introduces deep multi-task learning architectures with shared external memory, enabling joint training on related tasks to improve performance, especially when individual task data is limited.

Contribution

The paper proposes novel deep architectures that incorporate shared external memory for multi-task learning, enhancing performance across related text classification tasks.

Findings

Shared memory improves task performance

Joint training benefits individual tasks

Effective on multiple text classification tasks

Abstract

Neural network based models have achieved impressive results on various specific tasks. However, in previous works, most models are learned separately based on single-task supervised objectives, which often suffer from insufficient training data. In this paper, we propose two deep architectures which can be trained jointly on multiple related tasks. More specifically, we augment neural model with an external memory, which is shared by several tasks. Experiments on two groups of text classification tasks show that our proposed architectures can improve the performance of a task with the help of other related tasks.