We provide the first solution for model-free reinforcement learning of ω-regular objectives for Markov decision processes (MDPs). We present a constructive reduction from the almost-sure satisfaction of ω-regular objectives to an almostsure reachability problem, and extend this technique to learning how to control an unknown model so that the chance of satisfying the objective is maximized. A key feature of our technique is the compilation of ω-regular properties into limitdeterministic Büchi automata instead of the traditional Rabin automata; this choice sidesteps difficulties that have marred previous proposals. Our approach allows us to apply model-free, off-the-shelf reinforcement learning algorithms to compute optimal strategies from the observations of the MDP. We present an experimental evaluation of our technique on benchmark learning problems.An ω-word w on an alphabet Σ is a function w : N → Σ. We abbreviate w(i) by w i . The set of ω-words on Σ is written Σ ω and a subset of Σ ω is an ω-language on Σ.A probability distribution over a finite set S is a function d : S→[0, 1] such that s∈S d(s) = 1. Let D(S) denote the set of all discrete distributions over S. We say a distribution d ∈ D(S) is a point distribution if d(s)=1 for some s ∈ S. For a distribution d ∈ D(S) we write supp(d) def = {s ∈ S : d(s) > 0}.

We characterize the class of nondeterministic ω-automata that can be used for the analysis of finite Markov decision processes (MDPs). We call these automata 'good-for-MDPs' (GFM). We show that GFM automata are closed under classic simulation as well as under more powerful simulation relations that leverage properties of optimal control strategies for MDPs. This closure enables us to exploit state-space reduction techniques, such as those based on direct and delayed simulation, that guarantee simulation equivalence. We demonstrate the promise of GFM automata by defining a new class of automata with favorable properties-they are Büchi automata with low branching degree obtained through a simple construction-and show that going beyond limit-deterministic automata may significantly benefit reinforcement learning.

Omega-regular properties-specified using linear time temporal logic or various forms of omega-automata-find increasing use in specifying the objectives of reinforcement learning (RL). The key problem that arises is that of faithful and effective translation of the objective into a scalar reward for model-free RL. A recent approach exploits Büchi automata with restricted nondeterminism to reduce the search for an optimal policy for an ω-regular property to that for a simple reachability objective. A possible drawback of this translation is that reachability rewards are sparse, being reaped only at the end of each episode. Another approach reduces the search for an optimal policy to an optimization problem with two interdependent discount parameters. While this approach provides denser rewards than the reduction to reachability, it is not easily mapped to off-the-shelf RL algorithms. We propose a reward scheme that reduces the search for an optimal policy to an optimization problem with a single discount parameter that produces dense rewards and is compatible with off-the-shelf RL algorithms. Finally, we report an experimental comparison of these and other reward schemes for model-free RL with omega-regular objectives.

We study the problem of finding optimal strategies in Markov decision processes with lexicographic ω-regular objectives, which are ordered collections of ordinary ω-regular objectives. The goal is to compute strategies that maximise the probability of satisfaction of the first ω-regular objective; subject to that, the strategy should also maximise the probability of satisfaction of the second ω-regular objective; then the third and so forth. For instance, one may want to guarantee critical requirements first, functional ones second and only then focus on the non-functional ones. We show how to harness the classic off-the-shelf model-free reinforcement learning techniques to solve this problem and evaluate their performance on four case studies.