Research Technician at BCAM
OPTECOT - Optimal Evaluation Cost Tracking
This repository contains supplementary material for the paper Speeding-up Evolutionary Algorithms to solve Black-Box Optimization Problems. In this work, we have presented OPTECOT (Optimal Evaluation Cost Tracking): a technique to reduce the cost of solving a computationally expensive black-box optimization problem using population-based algorithms, avoiding loss of solution quality. OPTECOT requires a set of approximate objective functions of different costs and accuracies, obtained by modifying a strategic parameter in the definition of the original function. The proposal allows the selection of the lowest cost approximation with the trade-off between cost and accuracy in real time during the algorithm execution. To solve an optimization problem different from those addressed in the paper, the repository also contains a library to apply OPTECOT with the CMA-ES (Covariance Matrix Adaptation Evolution Strategy) optimization algorithm.
Authors: Judith Echevarrieta, Etor Arza, Aritz Pérez
License: free and open source software