1

Specification-Guided Reinforcement Learning
This tutorial explores specification-guided reinforcement learning as an alternative to traditional reward-based approaches, where the …
Kishor Jothimurugan, Suguman Bansal, Osbert Bastani, Rajeev Alur
Cite PDF
Robust Subtask Leaning for Compositional Generalization
Compositional reinforcement learning is a promising approach for training policies to perform complex long-horizon tasks. Typically, a …
Kishor Jothimurugan, Steve Hsu, Osbert Bastani, Rajeev Alur
Cite PDF ArXiv Code
Policy Synthesis and Reinforcement Learning for Discounted LTL
The difficulty of manually specifying reward functions has led to an interest in using linear temporal logic (LTL) to express …
Rajeev Alur, Osbert Bastani, Kishor Jothimurugan, Mateo Perez, Fabio Somenzi, Ashutosh Trivedi
Cite PDF ArXiv
Specification-Guided Learning of Nash Equilibria with High Social Welfare
Reinforcement learning has been shown to be an effective strategy for automatically training policies for challenging control problems. …
Kishor Jothimurugan, Suguman Bansal, Osbert Bastani, Rajeev Alur
Cite PDF ArXiv Code Slides
Compositional Reinforcement Learning from Logical Specifications
We study the problem of learning control policies for complex tasks given by logical specifications. Recent approaches automatically …
Kishor Jothimurugan, Suguman Bansal, Osbert Bastani, Rajeev Alur
Cite PDF ArXiv Code
Compositional Learning and Verification of Neural Network Controllers
Recent advances in deep learning have enabled data-driven controller design for autonomous systems. However, verifying safety of such …
Radoslav Ivanov, Kishor Jothimurugan, Steve Hsu, Shaan Vaidya, Rajeev Alur, Osbert Bastani
Cite PDF Code Slides
Abstract Value Iteration for Hierarchical Reinforcement Learning
We propose a novel hierarchical reinforcement learning framework for control with continuous state and action spaces. In our framework, …
Kishor Jothimurugan, Osbert Bastani, Rajeev Alur
Cite PDF ArXiv Code Slides Talk
Learning Algorithms for Regenerative Stopping Problems with Applications to Shipping Consolidation in Logistics
Abstract We study regenerative stopping problems in which the system starts anew whenever the controller decides to stop and the long-term average cost is to be minimized. Traditional model-based solutions involve estimating the underlying process from data and computing strategies for the estimated model.
Kishor Jothimurugan, Matthew Andrews, Jeongran Lee, Lorenzo Maggi
Sep 1, 2020
PDF ArXiv
Space-efficient Query Evaluation over Probabilistic Event Streams
Real-time decision making in IoT applications relies upon space-efficient evaluation of queries over streaming data. To model the …
Rajeev Alur, Yu Chen, Kishor Jothimurugan, Sanjeev Khanna
Cite PDF Slides
A Composable Specification Language for Reinforcement Learning Tasks
Reinforcement learning is a promising approach for learning control policies for robot tasks. However, specifying complex tasks (e.g., …
Kishor Jothimurugan, Rajeev Alur, Osbert Bastani
Cite PDF ArXiv Code Slides