1)

I am interested in completing research related to the use of machine learning to generate decision trees which control non-player character behavior in video games. Decision trees are relatively interpretable and have a high-level correspondence to behavior trees, which are used in the most common approach to AI for video games (Świechowski, 2022). I also enjoyed working with them when I took Artificial Intelligence and Machine Learning. The environment I would propose using for testing is the Micro-Game Karting template in the Unity Asset Store, which is available for free and was used for testing by Mas’udi (2021).

References

Chan, M. T., Chan, C. W., & Gelowitz, C. (2015). Development of a Car Racing Simulator Game Using Artificial Intelligence Techniques. International Journal of Computer Games Technology, 2015.

Guo, X., Singh, S., Lee, H., Lewis, R. L., & Wang, X. (2014). Deep Learning for Real-Time Atari Game Play Using Offline Monte-Carlo Tree Search Planning. Advances in Neural Information Processing Systems, 27.

Mas’udi, N. A., Jonemaro, E. M. A., Akbar, M. A., & Afirianto, T. (2021). Development of Non-Player Character for 3D Kart Racing Game Using Decision Tree. Fountain of Informatics Journal, 6(2), 51-60.

Świechowski, Maciej and Ślęzak, Dominik, Monte Carlo Tree Search as an Offline Training Data Generator for Decision-Tree Based Game Agents (2022). BDR-D-22-00241, Available at SSRN: https://ssrn.com/abstract=4152772 or http://dx.doi.org/10.2139/ssrn.4152772

2) 

I am considering completing research related to the use of databases in video games for purposes apart from data storage. While the article and dissertation by O’Grady (2019, 2021)  which are referenced below both appear to focus on the feasibility of that approach, it seems likely that additional exploration is warranted regarding demonstrable advantages. For this exploration, I would propose focusing on the execution of path-finding through a database management system, one of the main areas examined by O’Grady (2021).

References

Muhammad, Y. (2011). Evaluation and Implementation of Distributed NoSQL Database for MMO Gaming Environment (Dissertation, Uppsala University).

O’Grady, D. (2021). Bringing Database Management Systems and Video Game Engines Together (Doctoral dissertation, Eberhard Karls Universität Tübingen).

O’Grady, D. (2019). Database-Supported Video Game Engines: Data-Driven Map Generation. BTW 2019.

Jovanovic, R. (2013). Database Driven Multi-agent Behaviour Module (Thesis, York University).

3)

I am also considering completing research related to the use of machine learning for recognition of Kuzushiji, an old style of Japanese cursive writing. Kuzushiji mainly appear in works from the Edo period of Japanese history and are difficult to identify correctly due to their lack of standardization (Ueki, 2020). Another difficulty comes from the Chirashigaki writing style, in which text is not written in straight columns (Lamb, 2020). The Center for Open Data in the Humanities has released a dataset of them (Ueki, 2020), which is available at http://codh.rois.ac.jp/char-shape/. 

References

Clanuwat, T., Bober-Irizar, M., Kitamoto, A., Lamb, A., Yamamoto, K., & Ha, D. (2018). Deep Learning for Classical Japanese Literature. arXiv preprint arXiv:1812.01718.

Clanuwat, T., Lamb, A., & Kitamoto, A. (2019). KuroNet: Pre-Modern Japanese Kuzushiji Character Recognition with Deep Learning. arXiv preprint arXiv:1910.09433.

Ueki, K., & Kojima, T. (2020). Feasibility Study of Deep Learning Based Japanese Cursive Character Recognition. IIEEJ Transactions on Image Electronics and Visual Computing, 8(1), 10-16.