Learning Model
This program was developed to support research at Harvard University. I developed a simulation of the children's board game Funny Bunny to test Bayesian theories of human learning. The implementation was a naive predictor model that was used to compare to actual results of children playing the game. The program runs multiple iterations of the game and averages the number of turns spent to succesfully win the game. These results were then compared to the average number of turns by the children who played the board game.
The output is written to a .txt file that indicates each move and the subsequent weight applied to the move. Note that this can be ignored by setting a flag in the code so that only the end results are written out. The source code can be obtained below
download Predictor_Model