AdamPSO

A Particle Swarm Optimization Algorithm with Adaptive Moment Estimation.

Overview

AdamPSO is a Particle Swarm Optimization Algorithm with Adaptive Moment Estimation (Adam) method for single objective black-box optimization. In this approach, the learning rate in each dimension is independently adjusted in a self-adaptive manner. As a result, it improves the performance of the conventional PSO algorithm in some classic benchmarking functions.

The algorithm and the experimental results are detailed in [pdf].

Guideline for running the code

Run testEA.m for reproducing the experimental results.

Modify configurations.m if you want to customize the testing cases.

The implementation of AdamPSO is in optimisers/EA.m.

Reference

The testing framework is based on Dr. Liu and Mr. Pei's repo.