Forecasting the Direction of Short-Term Crude Oil Price Changes with Genetic-Fuzzy Information Distribution

Figure 1

The fitness values during optimization for forecasting WTI oil price (). The setting of genetic algorithm is as follows: the size of the population is 10; the fraction of the population at the next generation is 0.85; the fraction of individuals in each subpopulation that migrates to a different subpopulation is 0.2; and the optimization terminates when average change in the fitness value is less than 1e-6. The optimization process converged at the 100th generation. The estimated is with the best fitness value is 1.426.