Active Automata Learning from Noisy Data - Algorithms and Artifacts

This repository contains the benchmarking set, evaluation results and source code of the implementation for the Active Partial Max-SAT Learning (APMSL). It also contains the passive-to-active learning framework to make passive automata learning algorithms active together with all other artifacts presented in the paper "Active Automata Learning with Noisy Data: From Big to Small Data" by Felix Wallner, Bernhard Aichernig, Benjamin von Berg and Maximilian Rindler accepted to the FM 2026 Conference.

It also contains a pre-built multi-architecture docker image for ease of use.