Quantum-inspired evolutionary algorithms for financial data analysis

Abstract

This paper describes a real-valued quantum-inspired evolutionary algorithm (QIEA), a new computational approach which bears similarity with estimation of distribution algorithms (EDAs). The study assesses the performance of the QIEA on a series of benchmark problems and compares the results with those from a canonical genetic algorithm. Furthermore, we apply QIEA to a finance problem, namely non-linear principal component analysis of implied volatilities. The results from the algorithm are shown to be robust and they suggest potential for useful application of the QIEA to high-dimensional optimization problems in finance. © 2008 Springer-Verlag Berlin Heidelberg.