Publications
Ph.D. Dissertation
[link] Jesse Hostetler (2017) Monte Carlo Tree Search with Fixed and Adaptive Abstractions. PhD dissertation, Oregon State University.
Journal
[pdf] Jesse Hostetler, Alan Fern, and Thomas Dietterich (accepted). Monte Carlo tree search with fixed and adaptive state abstractions. Journal of AI Research (JAIR).
Conference
[pdf] Jesse Hostetler, Alan Fern, and Thomas Dietterich (2015). Progressive abstraction refinement in sparse sampling. Conference on Uncertainty in Artificial Intelligence (UAI).
[pdf] Jesse Hostetler, Alan Fern, and Thomas Dietterich (2014). State abstraction in Monte Carlo tree search. AAAI Conference on Artificial Intelligence.
[pdf] Brian King, Alan Fern, and Jesse Hostetler (2013). On adversarial policy switching with experiments in real-time strategy games. International Conference on Automated Planning and Scheduling (ICAPS).
[pdf] Jesse Hostetler, Ethan Derezynski, Thomas Dietterich, and Alan Fern (2012). Inferring strategies from limited reconnaissance in real-time strategy games. Conference on Uncertainty in Artificial Intelligence (UAI).
[pdf] Ethan Derezynski, Jesse Hostetler, Alan Fern, Thomas Dietterich, Thao-Trang Hoang, and Mark Udarbe (2011). Learning probabilistic behavior models in real-time strategy games. AAAI Conference on AI in Interactive Digital Entertainment (AIIDE).
G. Nugent, K. Kupzyk, S. A. Riley, L. D. Miller, J. Hostetler, L. K. Soh, and A. Samal (2009). Empirical usage metadata in learning objects. IEEE Fronties in Education Conference (FIE).
Workshop
[pdf] Brian King, Alan Fern, and Jesse Hostetler (2012). Adversarial policy switching with application to RTS games. AAAI Conference on AI in Interactive Digital Entertainment (AIIDE); Workshop on AI in Adversarial Real-Time Games.
Unrefereed
David Kortenkamp, Peter Bonasso, David Musliner, Michael Pelican, and Jesse Hostetler (2011). Embedding planning technology into satellite systems. AIAA Infotech@Aerospace.
Software
Progressive Abstraction Refinement for Sparse Sampling
A Java implementation of abstract Forward Search Sparse Sampling and the PARSS algorithm.
Note that the Github repository contains a lot of code related to other projects as well. The UAI snapshot may be your best option if just want to use or modify the PARSS algorithm.
Please reference the following paper in relation to the PARSS algorithm:
@inproceedings{hostetler2015progressive,
title={Progressive Abstraction Refinement for Sparse Sampling},
author={Hostetler, Jesse and Fern, Alan and Dietterich, Thomas},
booktitle={Conference on Uncertainty in Artificial Intelligence (UAI)},
year={2015}
}
Data
Starcraft: Broodwar Opening Strategy Dataset
ZIP archive (61 MB)
This dataset contains log data from 509 expert-level Starcraft: Broodwar games. All of the games are Protoss versus Terran matchups, and the dataset focuses on predicting the Protoss player's choices from the Terran player's perspective.
The games are divided into 30-second time steps, and the first 14 time steps from each game are retained. For each game, the data includes
- The total number of units of each type that were observable in each time step
- How many units of each type were produced
- How many units of each type were seen
- The Terran player's "scouting effort", which is the percentage of the Protoss player's Main Base + Natural Expansion that the Terran player could see
The README file included with the dataset explains the contents of each file in more detail.
Please reference the following paper in regards to this data:
@inproceedings{hostetler2012inferring,
title={Inferring Strategies from Limited Reconnaissance in Real-time Strategy Games},
author={Hostetler, Jesse and Derezynski, Ethan and Dietterich, Thomas and Fern, Alan},
booktitle={Conference on Uncertainty in Artificial Intelligence (UAI)},
year={2012}
}
