5 papers
Synthesizing POMDP Policies: Sampling Meets Model-checking via Learning
Debraj Chakraborty, Anirban Majumdar, Prince Mathew +2
Partially Observable Markov Decision Processes (POMDPs) are the standard framework for decision-making under uncertainty. While sampling-based methods scale well, they lack formal…
About Time: Model-free Reinforcement Learning with Timed Reward Machines
Rajarshi Roy, Anirban Majumdar, Ritam Raha +2
Reward specification plays a central role in reinforcement learning (RL), guiding the agent's behavior. To express non-Markovian rewards, formalisms such as reward machines have be…
Scalable Learning of One-Counter Automata via State-Merging Algorithms
Shibashis Guha, Anirban Majumdar, Prince Mathew +1
We propose One-counter Positive Negative Inference (OPNI), a passive learning algorithm for deterministic real-time one-counter automata (DROCA). Inspired by the RPNI algorithm for…
Learning Event-recording Automata Passively
Anirban Majumdar, Sayan Mukherjee, Jean-François Raskin
This paper presents a state-merging algorithm for learning timed languages definable by Event-Recording Automata (ERA) using positive and negative samples in the form of symbolic t…
Greybox Learning of Languages Recognizable by Event-Recording Automata
Anirban Majumdar, Sayan Mukherjee, Jean-François Raskin
In this paper, we revisit the active learning of timed languages recognizable by event-recording automata. Our framework employs a method known as greybox learning, which enables t…