Stochastic global optimization as a filtering problem

TL;DR

This paper reformulates stochastic global optimization as a filtering problem using particle filters, enabling faster convergence in noisy environments compared to naive methods.

Contribution

It introduces a novel filtering-based approach to stochastic global optimization, leveraging particle filters for improved convergence speed in noisy settings.

Findings

The filtering reformulation improves convergence speed.

Particle filters effectively focus on promising solutions.

The approach outperforms naive optimization in examples.

Abstract

We present a reformulation of stochastic global optimization as a filtering problem. The motivation behind this reformulation comes from the fact that for many optimization problems we cannot evaluate exactly the objective function to be optimized. Similarly, we may not be able to evaluate exactly the functions involved in iterative optimization algorithms. For example, we may only have access to noisy measurements of the functions or statistical estimates provided through Monte Carlo sampling. This makes iterative optimization algorithms behave like stochastic maps. Naive global optimization amounts to evolving a collection of realizations of this stochastic map and picking the realization with the best properties. This motivates the use of filtering techniques to allow focusing on realizations that are more promising than others. In particular, we present a filtering reformulation of global optimization in terms of a special case of sequential importance sampling methods called particle filters. The increasing popularity of particle filters is based on the simplicity of their implementation and their flexibility. For parametric exponential density estimation problems, we utilize the flexibility of particle filters to construct a stochastic global optimization algorithm which converges to the optimal solution appreciably faster than naive global optimization. Several examples are provided to demonstrate the efficiency of the approach.