The Tsetlin Machine Goes Deep: Logical Learning and Reasoning With Graphs

TL;DR

The paper introduces the Graph Tsetlin Machine (GraphTM), a novel interpretable deep clause learning model for graph-structured data, achieving high accuracy across diverse applications with improved efficiency and interpretability.

Contribution

It extends the Tsetlin Machine to handle graph-structured data using message passing, enabling deep clause learning for complex patterns with fewer clauses and better interpretability.

Findings

GraphTM achieves 3.86%-points higher accuracy on CIFAR-10 than convolutional TM.

Outperforms other reinforcement learning methods by up to 20.6%-points in action coreference tasks.

Tolerates significant noise in recommendation systems, similar to GCN, and trains 2.5x faster than GCN on viral genome data.

Abstract

Pattern recognition with concise and flat AND-rules makes the Tsetlin Machine (TM) both interpretable and efficient, while the power of Tsetlin automata enables accuracy comparable to deep learning on an increasing number of datasets. We introduce the Graph Tsetlin Machine (GraphTM) for learning interpretable deep clauses from graph-structured input. Moving beyond flat, fixed-length input, the GraphTM gets more versatile, supporting sequences, grids, relations, and multimodality. Through message passing, the GraphTM builds nested deep clauses to recognize sub-graph patterns with exponentially fewer clauses, increasing both interpretability and data utilization. For image classification, GraphTM preserves interpretability and achieves 3.86%-points higher accuracy on CIFAR-10 than a convolutional TM. For tracking action coreference, faced with increasingly challenging tasks, GraphTM outperforms other reinforcement learning methods by up to 20.6%-points. In recommendation systems, it tolerates increasing noise to a great extent, similar to a GCN. Finally, for viral genome sequence data, GraphTM is competitive with BiLSTM-CNN and GCN accuracy-wise, training ~2.5x faster than GCN. The GraphTM's application to these varied fields demonstrates how graph representation learning and deep clauses bring new possibilities for TM learning.