International Journal of Swarm Intelligence and Evolutionary Computation

Abstract

Topological Machine Learning and Chaotic Attractors Decomposition?An Application to Sunspot Chaos

Carlos Pedro Gonçalves*

We apply new empirical methods from chaos theory, aimed at dealing with stochastic chaos, employing adaptive topological artificial intelligence and topological data analysis to sunspots’ data for attractor reconstruction analysis and dynamical process decomposition. Applying these methods to sunspots’ data we uncover not one but two low-dimensional chaotic attractors, a first dominant attractor that is linked to a strongly persistent process with self-organized criticality and multifractal signatures, and a second chaotic attractor that exhibits intermittent turbulence and anti-persistent multifractal signatures, also present is a third process with an autoregressive moving average structure and an Independent and Identically Distributed (IID) noise component. In this way, the main emergent dynamics associated with the sunspots’ data are researched in detail down to the IID noise component.

Published Date: 2024-09-05; Received Date: 2024-08-06