Posts by Collection

portfolio

publications

Resource Constrained Deep Reinforcement Learning

Published in Proceedings of the International Conference on Automated Planning and Scheduling, 2019

TL;DR: Deep RL to optimize constrained resource allocation at city scale. Good results on realistic datasets. Read more

Recommended citation: Bhatia, A., Varakantham, P., & Kumar, A. (2019). Resource Constrained Deep Reinforcement Learning. In Proceedings of the International Conference on Automated Planning and Scheduling, 29(1), 610-620. https://ojs.aaai.org/index.php/ICAPS/article/view/3528

Tuning the Hyperparameters of Anytime Planning: A Deep Reinforcement Learning Approach

Published in ICAPS 2021 Workshop on Heuristics and Search for Domain-independent Planning, 2021

TL;DR: Deep RL to control hyperparameters of anytime algorithms at runtime to optimize quality of the final solution. Good results on Anytime A* search algorithm. Read more

Recommended citation: Bhatia, A., Svegliato, J., & Zilberstein, S. (2021). Tuning the Hyperparameters of Anytime Planning: A Deep Reinforcement Learning Approach. In ICAPS 2021 Workshop on Heuristics and Search for Domain-independent Planning. https://openreview.net/forum?id=c7hpFp_eRCo

On the Benefits of Randomly Adjusting Anytime Weighted A*

Published in Proceedings of the International Symposium on Combinatorial Search, 2021

TL;DR: Randomized Weighted A* tunes the weight in Anytime Weighted A* randomly at runtime and outperforms every static weighted baseline. Read more

Recommended citation: Bhatia, A., Svegliato, J., & Zilberstein, S. (2021). On the Benefits of Randomly Adjusting Anytime Weighted A. In Proceedings of the International Symposium on Combinatorial Search (Vol. 12, No. 1, pp. 116-120). https://ojs.aaai.org/index.php/SOCS/article/view/18558

Adaptive Rollout Length for Model-Based RL Using Model-Free Deep RL

Published in arXiv preprint arXiv:2206.02380, 2022

TL;DR: Meta-level deep RL to adapt the rollout-length in model-based RL non-myopically based on feedback from the learning process, such as accuracy of the model, learning progress and scarcity of samples. Read more

Recommended citation: Bhatia, A., Thomas, PS., & Zilberstein, S. (2022). Adaptive Rollout Length for Model-Based RL Using Model-Free Deep RL. In arXiv preprint arXiv:2206.02380. https://arxiv.org/abs/2206.02380

Tuning the Hyperparameters of Anytime Planning: A Metareasoning Approach with Deep Reinforcement Learning

Published in Proceedings of the International Conference on Automated Planning and Scheduling, 2022

TL;DR: Deep RL to determine optimal stopping point and hyperparameters of anytime algorithms at runtime to optimize utility of the final solution. Good results on Anytime A* search algorithm and RRT* motion planning algorithm. Read more

Recommended citation: Bhatia, A., Svegliato, J., Nashed, S. B., & Zilberstein, S. (2022). Tuning the Hyperparameters of Anytime Planning: A Metareasoning Approach with Deep Reinforcement Learning. In Proceedings of the International Conference on Automated Planning and Scheduling, 32(1), 556-564. https://ojs.aaai.org/index.php/ICAPS/article/view/19842

talks

teaching

Teaching experience 1

Undergraduate course, University 1, Department, 2014

This is a description of a teaching experience. You can use markdown like any other post. Read more

Teaching experience 2

Workshop, University 1, Department, 2015

This is a description of a teaching experience. You can use markdown like any other post. Read more