A Comparison of GA and PSO for Excess Return Evaluation in Stock Markets

摘要

One of the important problems in financial markets is making the profitable stocks trading rules using historical stocks market data. This paper implemented Particle Swarm Optimization (PSO) which is a new robust stochastic evolutionary computation Algorithm based on the movement and intelligence of swarms, and compared it to a Genetic Algorithm (GA) for generating trading rules. The results showed that PSO shares the ability of genetic algorithm to handle arbitrary nonlinear functions, but with a much simpler implementation clearly demonstrates good possibilities for use in Finance.