Interpreting Finite Automata for Sequential Data

TL;DR

This paper explores how to interpret finite automata for sequential data by identifying key interpretability properties and modifying state-merging algorithms, demonstrating their application in various prediction and clustering tasks.

Contribution

It introduces a modified state-merging approach for learning interpretable finite automata and applies it to diverse sequential data analysis scenarios.

Findings

Key properties for automaton interpretability are identified.

Modified automaton learning approach is effective across tasks.

Automata can be used for prediction, classification, and clustering.

Abstract

Automaton models are often seen as interpretable models. Interpretability itself is not well defined: it remains unclear what interpretability means without first explicitly specifying objectives or desired attributes. In this paper, we identify the key properties used to interpret automata and propose a modification of a state-merging approach to learn variants of finite state automata. We apply the approach to problems beyond typical grammar inference tasks. Additionally, we cover several use-cases for prediction, classification, and clustering on sequential data in both supervised and unsupervised scenarios to show how the identified key properties are applicable in a wide range of contexts.