In regular inference, the problem is to infer a regular language, typically represented by a deterministic finite automaton (DFA) from answers to a finite set of membership queries, each of which asks whether the language contains a certain word. There are many algorithms for learning DFAs, the most well-known being the L* algorithm due to Dana Angluin. However, there are almost no extensions of these algorithms to the setting of timed systems. We present an algorithm for inferring a model of a timed system using Angluin's setup. One of the most popular model for timed system is timed automata. Since timed automata can freely use an arbitrary number of clocks, we restrict our attention to systems that can be described by event-recording automata (DERAs). In previous work, we have presented an algorithm for inferring a DERA in the form of a region graph. In this paper, we present a novel inference algorithm for DERAs, which avoids constructing a (usually prohibitively large) region graph. We must then develop techniques for inferring guards on transitions of a DERA. Our construction deviates from previous work on inference of DERAs in that it first constructs a so called timed decision tree from observations of system behavior, which is thereafter folded into an automaton.
Download BibTeX entry.