Creating Realistic AI Using N-Grams
Presentation explaining what N-Grams are, why to use them, optimization techniques, and a demonstration of N-Grams in action.
The first part of the demo begins at 5:31 in the video, and the second part begins at 8:32.
Project Details
Project Takeaways
Role and Project Dates:
Researcher
11/2021
Concept:
This research project was focused on N-Grams and the different optimizations that could be applied to better predict player behavior. Specifically, we experimented with finding an optimal N value for our combat simulation and explored N-Grams that were supported by recorded player reactions which we called player-generated N-Grams.
Languages and Tools Used:
Unity
C# scripts
Team Breakdown:
Programmers / Researchers (2)
Learned and applied methods to generate an engaging presentation, in order to showcase our research and provide a walkthrough of our demo.
Created a professional research paper detailing what we had learned during our research in order to explain the basic principles of N-Grams and document our findings about the usability of player-generated N-Grams.
Created a demo in Unity to showcase the power of using N-Grams to control an AI enemy in a simple fighting simulator. The demo also provided a method to demonstrate optimization methods for N-Grams and an example of how to create player-generated N-Grams.
Learned about the various methods and reasons for applying N-Grams to games by applying the concepts to an AI agent within a demo and expanding the basic N-Gram to account for player profiles in the form of player-generated N-Grams.
Project Download