Genetic programming (GP) is a general problem solving approach that uses evolutionary dynamics to find computer programs that solve the specified problems when executed. GP has been applied to a wide range of problems in various domains; in particular, for solving problems in quantitative finance and in econometrics. In this chapter we describe the fundamentals of GP and evolutionary algorithms and give a brief survey of relevant literature and results that have been achieved using GP in financial applications. We also present in detail how GP can be applied for identification of variable interaction networks and for prediction of multivariate nonlinear time-series. Furthermore, we demonstrate these two approaches using two practical examples, namely the identification of dependencies of leading indicators and economic variables in the US economy, and the prediction of European interest rate swap rates.
|Title of host publication||Recent Advances in Computational Finance|
|Publisher||Nova Science Publishers, Inc.|
|Number of pages||17|
|Publication status||Published - 2013|