Applications of genetic algorithms

In many heuristic optimization, it is easy to be trapped in local optimal. In contrast, genetic algorithms work from a population of solutions simultaneously, climbing many peaks in parallel. Thus, the probability of finding a false peak(local optimal) is greatly reduced. In our study, we apply genetic algorithm in classification problem and improve the performance of TES tool. In order to improve the performance of genetic algorithm and to solve the early convergence problem, we apply the idea of simulated annealing technique and test the annealed genetic algorithm on financial series, DNA sequences, as well as correlated time series with long and short memory. The result indicated that the annealed algorithm is much better than traditional ones