Team Number: 053 A
Area of Science: Artificial Intelligence
Project Title: Tic Tac Toe Time
Problem Definition: The problem that this project addresses is how to program a game to give a “learned” response. Teaching computers how to give various responses based on input from their environment, has many uses in the world today. Games are the most common place where people are trying to implement “learned” responses. One example of using “learned” responses would be Neural Networks; they have many different ways of sorting out the “learned” data from each input.
Goal: The goal of this project is to create a game program that will be able to play tic tac toe and learn from each game. We will use Neural Networks to sort the data of each game. Throughout the time of the game it should become increasingly harder for the player because the computer will have learned all of the players possible moves, from the previous games. After a certain amount of time, the game will always end up as a tie.
Problem Solution: The way that we would be able to solve our problem would have to be using Neural Networks “Learning Process.” Neural Nets has two different ways that we could approach the problem. The first choice that we have would be associative mapping, in which the network stores the relationship among patterns. Our other choice would be have to be regularity detection. This choice takes the past input information and applies it to each new event. We believe that the regularity detection would be the best choice for our programming because it has all of the programming types that we need to accomplish our goal.
Progress to Date: The progress that we have made so far has mainly been research. We now know that we will be using the Java system to program our game. We hope to be able to use Neural Networks to assist us on our attempt to create a Tic Tac Toe game.
Expected Results: We hope to have our complete program finished by the end of the challenge. We are also hoping that anyone will be able to play this game, not only programmers. The game will only last for a limited amount of time, though. The computer will eventually learn all possible moves and defeat the player or end in a tie for each game.
Team Members: Felicia Picknell
Sponsoring Teachers: Debra Loftin
http://www.doc.ic.ac.uk/~nd/surprise_96/journal/vol4/cs11/report.html#The Learning Process