DDNAS: Discretized Differentiable Neural Architecture Search for Text Classification

TL;DR

DDNAS introduces a novel differentiable neural architecture search method for text classification that optimizes architecture via gradient descent and models hierarchical categorization through a discretization layer, outperforming existing NAS approaches.

Contribution

The paper proposes DDNAS, a new NAS framework for text classification that combines continuous relaxation with a discretization layer based on mutual information maximization.

Findings

DDNAS outperforms state-of-the-art NAS methods on eight datasets.

Uses only three basic operations: convolution, pooling, and none.

Performance can be further improved by adding more operations.

Abstract

Neural Architecture Search (NAS) has shown promising capability in learning text representation. However, existing text-based NAS neither performs a learnable fusion of neural operations to optimize the architecture, nor encodes the latent hierarchical categorization behind text input. This paper presents a novel NAS method, Discretized Differentiable Neural Architecture Search (DDNAS), for text representation learning and classification. With the continuous relaxation of architecture representation, DDNAS can use gradient descent to optimize the search. We also propose a novel discretization layer via mutual information maximization, which is imposed on every search node to model the latent hierarchical categorization in text representation. Extensive experiments conducted on eight diverse real datasets exhibit that DDNAS can consistently outperform the state-of-the-art NAS methods. While DDNAS relies on only three basic operations, i.e., convolution, pooling, and none, to be the candidates of NAS building blocks, its promising performance is noticeable and extensible to obtain further improvement by adding more different operations.