diff --git a/content/paper/2403.01.md b/content/paper/2403.01.md new file mode 100644 index 0000000..972961f --- /dev/null +++ b/content/paper/2403.01.md @@ -0,0 +1,25 @@ +--- +draft: false +title: "Partially Observable Hierarchical Reinforcement Learning with AI Planning (Student Abstract)" +authors: [ + "Brandon Rozek", + "Junkyu Lee", + "Harsha Kokel", + "Michael Katz", + "Shirin Sohrabi" +] +date: 2024-03-24 +publish_date: "2024/03/24" +conference: "AAAI Conference on Artificial Intelligence" + + +isbn: "" +doi: "10.1609/aaai.v38i21.30504" +volume: 38 +firstpage: 23635 +lastpage: 23636 +language: "English" + +pdf_url: "https://ojs.aaai.org/index.php/AAAI/article/view/30504/32640" +abstract: "Partially observable Markov decision processes (POMDPs) challenge reinforcement learning agents due to incomplete knowledge of the environment. Even assuming monotonicity in uncertainty, it is difficult for an agent to know how and when to stop exploring for a given task. In this abstract, we discuss how to use hierarchical reinforcement learning (HRL) and AI Planning (AIP) to improve exploration when the agent knows possible valuations of unknown predicates and how to discover them. By encoding the uncertainty in an abstract planning model, the agent can derive a high-level plan which is then used to decompose the overall POMDP into a tree of semi-POMDPs for training. We evaluate our agent's performance on the MiniGrid domain and show how guided exploration may improve agent performance." +--- \ No newline at end of file diff --git a/content/publications.md b/content/publications.md index b3a63a1..3b6c8d8 100644 --- a/content/publications.md +++ b/content/publications.md @@ -8,10 +8,10 @@ aliases: ## Publications -Partially Observable Hierarchical Reinforcement Learning with AI Planning (Student Abstract) +[Partially Observable Hierarchical Reinforcement Learning with AI Planning (Student Abstract)](/paper/2403.01) - Authors: *Brandon Rozek*, Junkyu Lee, Harsha Kokel, Michael Katz and Shirin Sohrabi - Venue: AAAI Conference on Artificial Intelligence, 2024. -- Paper to appear in early 2024 | Poster to appear in early 2024 +- [Paper](https://ojs.aaai.org/index.php/AAAI/article/view/30504/32640) | [Poster](/files/research/AAAI_Remote_Poster_Rozek24.pdf) [Parallel Verification of Natural Deduction Proof Graphs](/paper/2311.01/) - Authors: James Oswald and *Brandon Rozek* @@ -61,7 +61,7 @@ Verification of Automatically Synthesized Cryptosystems ](/paper/2109.01/) Efficient Parallel Verification of Natural Deduction Proof Graphs - Authors: James Oswald and *Brandon Rozek* -- Venue: Rensselaer Computer Science Graduate Poster Session, April 2019. +- Venue: Rensselaer Computer Science Graduate Poster Session, April 2023. - [Poster](/files/research/PV-Poster.pdf) diff --git a/static/files/research/AAAI_Remote_Poster_Rozek24.pdf b/static/files/research/AAAI_Remote_Poster_Rozek24.pdf new file mode 100644 index 0000000..e47926d Binary files /dev/null and b/static/files/research/AAAI_Remote_Poster_Rozek24.pdf differ