Forecast of industrial index based on genetic neural network

In order to make more accurate forecast results, we combine methods of genetic algorithm with BNPP in the forecast of industrial index, otherwise in the process of the forecast we use five input-variables namely weekly turnover increase and decrease piece, weekly oscillation amplitude, weekly UP/DOWN, five bargain days' EMA fluctuation, DIF fluctuation and one output-variable The Next week UP/DOWN. As to the problem of low efficiency in choosing appropriate topological structure of neural network, the introduction of genetic algorithm has improved the speed of searching for the optimum topological structure greatly in this paper.