Machine Learning with Chaotic Strange Attractors

TL;DR

This paper introduces an analog computing approach using chaotic nonlinear attractors for machine learning, achieving high accuracy and low power consumption, addressing the limitations of traditional digital methods.

Contribution

The authors propose a novel analog computing platform based on chaotic attractors that is programmable, versatile, and energy-efficient for machine learning tasks.

Findings

Achieves high accuracy in regression and classification tasks.

Operates at milliwatt-scale power levels, comparable to current methods.

Provides low error rates in machine learning applications.

Abstract

Machine learning studies need colossal power to process massive datasets and train neural networks to reach high accuracies, which have become gradually unsustainable. Limited by the von Neumann bottleneck, current computing architectures and methods fuel this high power consumption. Here, we present an analog computing method that harnesses chaotic nonlinear attractors to perform machine learning tasks with low power consumption. Inspired by neuromorphic computing, our model is a programmable, versatile, and generalized platform for machine learning tasks. Our mode provides exceptional performance in clustering by utilizing chaotic attractors' nonlinear mapping and sensitivity to initial conditions. When deployed as a simple analog device, it only requires milliwatt-scale power levels while being on par with current machine learning techniques. We demonstrate low errors and high accuracies with our model for regression and classification-based learning tasks.