Abstract
Recent experiments have shown that a deep neural network can be trained to predict the action of steps of Conway's Game of Life automaton given millions of examples of this action on random initial states. However, training was never completely successful for , and even when successful, a reconstruction of the elementary rule () from data is not within the scope of what the neural network can deliver. We describe an alternative network-like method, based on constraint projections, where this is possible. From a single data item this method perfectly reconstructs not just the automaton rule but also the states in the time steps it did not see. For a unique reconstruction, the size of the initial state need only be large enough that it and the states it evolves into contain all possible automaton input patterns. We demonstrate the method on 1D binary cellular automata that take inputs from adjacent cells. The unknown rules in our experiments are not restricted to simple rules derived from a few linear functions on the inputs (as in Game of Life), but include all possible rules on inputs. Our results extend to , for which exhaustive rule-search is not feasible. By relaxing translational symmetry in space and also time, our method is attractive as a platform for the learning of binary data, since the discreteness of the variables does not pose the same challenge it does for gradient-based methods.
8 More- Received 7 December 2020
- Revised 9 June 2021
- Accepted 17 August 2021
DOI:https://doi.org/10.1103/PhysRevE.104.034301
©2021 American Physical Society