ECCO: Equivalent Circuit Controlled Optimization

TL;DR

This paper introduces ECCO, an adaptive optimization algorithm inspired by electrical circuit dynamics, which controls the optimization trajectory and step sizes to achieve fast convergence on smooth functions, outperforming traditional methods like Adam.

Contribution

The paper presents a novel optimization approach that leverages circuit theory to control gradient flow, including new trajectory control schemes and an error-aware discretization method.

Findings

Outperforms uncontrolled gradient flow in simulations.

Error-aware discretization surpasses Armijo line search.

Achieves convergence speeds comparable to or better than Adam.

Abstract

We propose an adaptive optimization algorithm for solving unconstrained scaled gradient flow problems that achieves fast convergence by controlling the optimization trajectory shape and the discretization step sizes. Under a broad class of scaling functions, we establish convergence of the proposed approach to critical points of smooth objective functions, while demonstrating its flexibility and robustness with respect to hyperparameter tuning. First, we prove convergence of component-wise scaled gradient flow to a critical point under regularity conditions. We show that this controlled gradient flow dynamics is equivalent to the transient response of an electrical circuit, allowing for circuit theory concepts to solve the problem. Based on this equivalence, we develop two optimization trajectory control schemes based on minimizing the charge stored in the circuit: a second order method that uses the true Hessian and an alternate first order method that approximates the optimization trajectory with only gradient information. While the control schemes are derived from circuit concepts, no circuit knowledge is needed to implement the algorithms. To find the value of the critical point, we propose a time step search routine for Forward Euler discretization that controls the local truncation error, a method adapted from circuit simulation ideas. In simulation we find that the trajectory control outperforms uncontrolled gradient flow, and the error-aware discretization out-performs line search with the Armijo condition. Our algorithms are evaluated on convex and non-convex test functions, including neural networks, with convergence speeds comparable to or exceeding Adam.