Data-Centric Approach to Constrained Machine Learning: A Case Study on Conway's Game of Life

TL;DR

This study demonstrates that a carefully designed training dataset, combined with domain expertise, significantly improves the performance of minimal neural networks in learning Conway's Game of Life transition rules under parameter constraints.

Contribution

It introduces a data-centric methodology emphasizing dataset design and domain knowledge for constrained machine learning tasks, exemplified through Conway's Game of Life.

Findings

Strategic dataset design enhances learning outcomes.

Domain expert insights are crucial for effective training.

Benefits of dataset strategy persist across different training configurations.

Abstract

This paper focuses on a data-centric approach to machine learning applications in the context of Conway's Game of Life. Specifically, we consider the task of training a minimal architecture network to learn the transition rules of Game of Life for a given number of steps ahead, which is known to be challenging due to restrictions on the allowed number of trainable parameters. An extensive quantitative analysis showcases the benefits of utilizing a strategically designed training dataset, with its advantages persisting regardless of other parameters of the learning configuration, such as network initialization weights or optimization algorithm. Importantly, our findings highlight the integral role of domain expert insights in creating effective machine learning applications for constrained real-world scenarios.