Feasability of Learning Weighted Automata on a Semiring

TL;DR

This paper investigates the feasibility and limitations of learning weighted automata over semirings using Angluin-style algorithms, providing a theoretical classification of learnability and discussing algorithmic challenges.

Contribution

It introduces a theoretical framework classifying functions based on their learnability by weighted automata and analyzes the limitations of Angluin's approach in this context.

Findings

Classifies functions by their guessability and automaton construction

Identifies theoretical limitations of Angluin-style learning for weighted automata

Discusses conditions for algorithmic solutions over various semirings

Abstract

Since the seminal work by Angluin and the introduction of the L*-algorithm, active learning of automata by membership and equivalence queries has been extensively studied to learn various extensions of automata. For weighted automata, algorithms for restricted cases have been developed in the literature, but so far there was no global approach or understanding how these algorithms could apply (or not) in the general case. In this paper we chart the boundaries of the Angluin approach. We use a class of hypothesis automata which are constructed, in Angluin's style, by using membership and equivalence queries and solving certain finite systems of linear equations over the semiring, and we show the theoretical limitations of this approach. We classify functions with respect to how guessable they are, corresponding to the existence of hypothesis automata computing a given function, and how such an hypothesis automaton can be found. Of course, from an algorithmic standpoint, knowing that a solution (hypothesis automaton) exists need not translate into an effective algorithm to find one. We relate our work to the existing literature with a discussion of some known properties ensuring algorithmic solutions, illustrating the ideas over several familiar semirings (including the natural numbers).