On when is Reservoir Computing with Cellular Automata Beneficial?

TL;DR

This paper investigates the effectiveness of Reservoir Computing with Cellular Automata (ReCA), highlighting its potential benefits, limitations, and the importance of task-specific suitability and ablation testing.

Contribution

It demonstrates that ReCA can be effective even in simple implementations and emphasizes the need for thorough testing to understand when ReCA is beneficial.

Findings

ReCA shows effectiveness in certain tasks due to encoding schemes.

ReCA's success may be primarily due to encoding rather than CA dynamics.

The paper raises questions about the types of tasks best suited for ReCA.

Abstract

Reservoir Computing with Cellular Automata (ReCA) is a relatively novel and promising approach. It consists of 3 steps: an encoding scheme to inject the problem into the CA, the CA iterations step itself and a simple classifying step, typically a linear classifier. This paper demonstrates that the ReCA concept is effective even in arguably the simplest implementation of a ReCA system. However, we also report a failed attempt on the UCR Time Series Classification Archive where ReCA seems to work, but only because of the encoding scheme itself, not in any part due to the CA. This highlights the need for ablation testing, i.e., comparing internally with sub-parts of one model, but also raises an open question on what kind of tasks ReCA is best suited for.