Conway’s Game of Life Replicated and Generalized
Project Manager: Nash Rickert
- Contact: firstname.lastname@example.org
The idea of the project would be to reimplement a prior paper which examined the difficulty of using a convolutional network to predict future steps of Conway’s Game of Life due to the Lottery Ticket Hypothesis (https://browse.arxiv.org/pdf/2009.01398.pdf). After reimplementation, which would serve as an opportunity to become more familiar with the programming aspect of machine learning, I would seek to generalize the findings to other cellular automata systems. In particular, the goal would be to see if the difficulty of predicting future states of a well-defined cellular automata is partially due to the discreteness of the conditions of Conway’s Game of Life that make it harder for the gradient to work effectively. It would be interesting to see if the same results do or do not generalize to a more ‘continuous’ scenario.
A large part of I2’s mission is to expand and further knowledge of machine learning. This experiment would allow me both a personal exploration of the topic and would hopefully yield a result useful to others.