Active Learning of Mealy Machines with Timers

2024-03-04

arXiv.org

AI Summary

This paper addresses the active learning of Timed Mealy Machines (TMMs) in black-box settings—a previously unexplored problem. Methodologically, it extends the classical L# algorithm to the timed domain by introducing symbolic queries for efficient inference of untimed behavior while supporting timed automaton modeling; it further proposes a concrete, implementable query interface that balances theoretical soundness with engineering feasibility. A Rust-based prototype is implemented and evaluated on real-world benchmarks, demonstrating rapid convergence, low query complexity, and practical effectiveness. Key contributions include: (i) the first black-box learning framework for TMMs; (ii) synergistic modeling of symbolic queries and timed semantics; and (iii) an industrially viable, efficient, and implementable solution tailored to real-world deployment constraints.

Technology Category

• Machine Learning: Active Learning
• Search and Optimization: Learning to Search
• Planning, Routing, and Scheduling: Learning for Planning and Scheduling

Application Category

• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions
• Search and Retrieval-Augmented AI: Search Tool Learning with LLM: Teaching LLMs to invoke search and make use of retrieved information
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications

Abstract

We present the first algorithm for query learning of a class of Mealy machines with timers in a black-box context. Our algorithm is an extension of the L# algorithm of Vaandrager et al. to a timed setting. We rely on symbolic queries which empower us to reason on untimed executions while learning. Similarly to the algorithm for learning timed automata of Waga, these symbolic queries can be implemented using finitely many concrete queries. Experiments with a prototype implementation, written in Rust, show that our algorithm is able to efficiently learn realistic benchmarks.

Problem

• Learning Mealy machines with timers algorithmically
• Extending L# algorithm to timed settings
• Using symbolic queries for untimed execution reasoning

Innovation

• Active learning algorithm for Mealy machines
• Extension of L# algorithm to timed setting
• Symbolic queries enable efficient timed learning