This paper addresses the problem of maintaining safety during training in Reinforcement Learning (RL), such that the safety constraint violations are bounded at any point during learning. As enforcing safety during training might severely limit the agent’s exploration, we propose here a new architecture that handles the trade-off between efficient progress and safety during exploration. As the exploration progresses, we update via Bayesian inference Dirichlet-Categorical models of the transition probabilities of the Markov decision process that describes the environment dynamics. We then propose a way to approximate moments of belief about the risk associated to the action selection policy. We demonstrate that this approach can be easily interleaved with RL and we present experimental results to showcase the performance of the overall architecture.

Safeguarded Progress in Reinforcement Learning: Safe Bayesian Exploration for Control Policy Synthesis | VIDEO

Neural-based, data-driven analysis and control of dynamical systems have been recently investigated and have shown great promise, e.g. for safety verification or stability analysis. Indeed, not only do neural networks allow for an entirely model-free, data-driven approach, but also for handling arbitrary complex functions via their power of representation (as opposed to, e.g. algebraic optimization techniques that are restricted to polynomial functions). Whilst classical Lyapunov techniques allow to provide a formal and robust guarantee of stability of a switched dynamical system, very little is yet known about correctness guarantees for Neural Lyapunov functions, nor about their performance (amount of data needed for a certain accuracy). 
We formally introduce Neural Lyapunov functions for the stability analysis of switched linear systems: we benchmark them on this paradigmatic problem, which is notoriously difficult (and in general Turing-undecidable), but which admits existing recently-developed technologies and theoretical results. Inspired by switched systems theory, we provide theoretical guarantees on the representative power of neural networks, leveraging recent results from the ML community. We additionally experimentally display how Neural Lyapunov functions compete with state-of-the-art results and techniques, while admitting a wide range of improvement, both in theory and in practice. This study intends to improve our understanding of the opportunities and current limitations of neural-based data-driven analysis and control of complex dynamical systems.

Stability Analysis of Switched Linear Systems with Neural Lyapunov Functions

The aim of Inverse Reinforcement Learning (IRL) is to infer a reward function R from a policy pi. To do this, we need a model of how pi relates to R. In the current literature, the most common models are optimality, Boltzmann rationality, and causal entropy maximisation. One of the primary motivations behind IRL is to infer human preferences from human behaviour. However, the true relationship between human preferences and human behaviour is much more complex than any of the models currently used in IRL. This means that they are misspecified, which raises the worry that they might lead to unsound inferences if applied to real-world data. In this paper, we provide a mathematical analysis of how robust different IRL models are to misspecification, and answer precisely how the demonstrator policy may differ from each of the standard models before that model leads to faulty inferences about the reward function R. We also introduce a framework for reasoning about misspecification in IRL, together with formal tools that can be used to easily derive the misspecification robustness of new IRL models.

Misspecification in Inverse Reinforcement Learning

Heating and cooling systems in buildings account for 31% of global energy use, much of which are regulated by Rule Based Controllers (RBCs) that neither maximise energy efficiency nor minimise emissions by interacting optimally with the grid. Control via Reinforcement Learning (RL) has been shown to significantly improve building energy efficiency, but existing solutions require access to building-specific simulators or data that cannot be expected for every building in the world. In response, we show it is possible to obtain emission-reducing policies without such knowledge a priori–a paradigm we call zero-shot building control. We combine ideas from system identification and model-based RL to create PEARL (Probabilistic Emission-Abating Reinforcement Learning) and show that a short period of active exploration is all that is required to build a performant model. In experiments across three varied building energy simulations, we show PEARL outperforms an existing RBC once, and popular RL baselines in all cases, reducing building emissions by as much as 31% whilst maintaining thermal comfort. Our source code is available online via: https://enjeeneer.io/projects/pearl/.

Low Emission Building Control with Zero-Shot Reinforcement Learning

Capturing uncertainty in models of complex dynamical systems is crucial to designing safe controllers. Stochastic noise causes aleatoric uncertainty, whereas imprecise knowledge of model parameters leads to epistemic uncertainty. Several approaches use formal abstractions to synthesize policies that satisfy temporal specifications related to safety and reachability. However, the underlying models exclusively capture aleatoric but not epistemic uncertainty, and thus require that model parameters are known precisely. Our contribution to overcoming this restriction is a novel abstraction-based controller synthesis method for continuous-state models with stochastic noise and uncertain parameters. By sampling techniques and robust analysis, we capture both aleatoric and epistemic uncertainty, with a user-specified confidence level, in the transition probability intervals of a so-called interval Markov decision process (iMDP). We synthesize an optimal policy on this iMDP, which translates (with the specified confidence level) to a feedback controller for the continuous model with the same performance guarantees. Our experimental benchmarks confirm that accounting for epistemic uncertainty leads to controllers that are more robust against variations in parameter values.

Probabilities Are Not Enough: Formal Controller Synthesis for Stochastic Dynamical Models with Epistemic Uncertainty

Controllers for autonomous systems that operate in safety-critical settings must account for stochastic disturbances. Such disturbances are often modeled as process noise, and common assumptions are that the underlying distributions are known and/or Gaussian. In practice, however, these assumptions may be unrealistic and can lead to poor approximations of the true noise distribution. We present a novel planning method that does not rely on any explicit representation of the noise distributions. In particular, we address the problem of computing a controller that provides probabilistic guarantees on safely reaching a target. First, we abstract the continuous system into a discrete-state model that captures noise by probabilistic transitions between states. As a key contribution, we adapt tools from the scenario approach to compute probably approximately correct (PAC) bounds on these transition probabilities, based on a finite number of samples of the noise. We capture these bounds in the transition probability intervals of a so-called interval Markov decision process (iMDP). This iMDP is robust against uncertainty in the transition probabilities, and the tightness of the probability intervals can be controlled through the number of samples. We use state-of-the-art verification techniques to provide guarantees on the iMDP, and compute a controller for which these guarantees carry over to the autonomous system. Realistic benchmarks show the practical applicability of our method, even when the iMDP has millions of states or transitions.

Sampling-Based Robust Control of Autonomous Systems with Non-Gaussian Noise

In this paper, we study the problem of learning to satisfy temporal logic specifications with a group of agents in an unknown environment, which may exhibit probabilistic behaviour. From a learning perspective these specifications provide a rich formal language with which to capture tasks or objectives, while from a logic and automated verification perspective the introduction of learning capabilities allows for practical applications in large, stochastic, unknown environments. The existing work in this area is, however, limited. Of the frameworks that consider full linear temporal logic or have correctness guarantees, all methods thus far consider only the case of a single temporal logic specification and a single agent. In order to overcome this limitation, we develop the first multi-agent reinforcement learning technique for temporal logic specifications, which is also novel in its ability to handle multiple specifications. We provide correctness and convergence guarantees for our main algorithm – ALMANAC (Automaton/Logic Multi-Agent Natural Actor-Critic) – even when using function approximation. Alongside our theoretical results, we further demonstrate the applicability of our technique via a set of preliminary experiments.

Multi-Agent Reinforcement Learning with Temporal Logic Specifications

Multi-agent influence diagrams (MAIDs) are a popular form of graphical model that, for certain classes of games, have been shown to offer key complexity and explainability advantages over traditional extensive form game (EFG) representations. In this paper, we extend previous work on MAIDs by introducing the concept of a MAID subgame, as well as subgame perfect and trembling hand perfect equilibrium refinements. We then prove several equivalence results between MAIDs and EFGs. Finally, we describe an open source implementation for reasoning about MAIDs and computing their equilibria.

Equilibrium Refinements for Multi-Agent Influence Diagrams: Theory and Practice

This paper proposes DeepSynth, a method for effective training of deep Reinforcement Learning (RL) agents when the reward is sparse and non-Markovian, but at the same time progress towards the reward requires achieving an unknown sequence of high-level objectives. Our method employs a novel algorithm for synthesis of compact automata to uncover this sequential structure automatically. We synthesise a human-interpretable automaton from trace data collected by exploring the environment. The state space of the environment is then enriched with the synthesised automaton so that the generation of a control policy by deep RL is guided by the discovered structure encoded in the automaton. The proposed approach is able to cope  with both high-dimensional, low-level features and unknown sparse non-Markovian rewards. We have evaluated DeepSynth's performance in a set of experiments that includes the Atari game Montezuma's Revenge. Compared to existing approaches, we obtain a reduction of two orders of magnitude in the number of iterations required for policy synthesis, and also a significant improvement in scalability.

DeepSynth: Automata Synthesis for Automatic Task Segmentation in Deep Reinforcement Learning

This work presents the concept of an adaptive safe padding that forces Reinforcement Learning (RL) to synthesise optimal control policies while ensuring safety during the learning process. Policies are synthesised to satisfy a goal, expressed as a temporal logic formula, with maximal probability. Enforcing the RL agent to stay safe during learning might limit the exploration, however we show that the proposed architecture is able to automatically handle the trade-off between efficient progress in exploration (towards goal satisfaction) and ensuring safety. Theoretical guarantees are available on the optimality of the synthesised policies and on the convergence of the learning algorithm. Experimental results are provided to showcase the performance of the proposed method.

Alessandro Abate

10

44

SHORT BIO