Output-decomposed Learning of Mealy Machines

TL;DR

This paper introduces an active automata learning algorithm that decomposes finite state machines based on output projections, reducing complexity and query count while enabling full system reconstruction.

Contribution

It presents a novel output-decomposed learning method for Mealy machines that is dual to existing approaches, improving efficiency through output projections.

Findings

Reduces number of queries needed for learning

Enables full system reconstruction from projections

Shows promising preliminary evaluation results

Abstract

We present an active automata learning algorithm which learns a decomposition of a finite state machine, based on projecting onto individual outputs. This is dual to a recent compositional learning algorithm by Labbaf et al. (2023). When projecting the outputs to a smaller set, the model itself is reduced in size. By having several such projections, we do not lose any information and the full system can be reconstructed. Depending on the structure of the system this reduces the number of queries drastically, as shown by a preliminary evaluation of the algorithm.