Recursive Tree Grammar Autoencoders

TL;DR

This paper introduces RTG-AE, a novel recursive neural network-based autoencoder that encodes and decodes trees using learned grammars, outperforming existing models in efficiency and accuracy for tree-structured data.

Contribution

It presents the first model combining variational autoencoders, grammatical knowledge, and recursive processing for tree data.

Findings

Improves autoencoding error and training time

Enhances optimization scores on datasets

Handles all regular tree grammars efficiently

Abstract

Machine learning on trees has been mostly focused on trees as input to algorithms. Much less research has investigated trees as output, which has many applications, such as molecule optimization for drug discovery, or hint generation for intelligent tutoring systems. In this work, we propose a novel autoencoder approach, called recursive tree grammar autoencoder (RTG-AE), which encodes trees via a bottom-up parser and decodes trees via a tree grammar, both learned via recursive neural networks that minimize the variational autoencoder loss. The resulting encoder and decoder can then be utilized in subsequent tasks, such as optimization and time series prediction. RTG-AEs are the first model to combine variational autoencoders, grammatical knowledge, and recursive processing. Our key message is that this unique combination of all three elements outperforms models which combine any two of the three. In particular, we perform an ablation study to show that our proposed method improves the autoencoding error, training time, and optimization score on synthetic as well as real datasets compared to four baselines. We further prove that RTG-AEs parse and generate trees in linear time and are expressive enough to handle all regular tree grammars.