The paper introduces a method of evolutionary computer programming using genetic algorithms to automatically establish a dynamic nonlinear mathematical model for data mining, and perform social and economic trend prediction and regression curve rounding, changing the past method of only using rough rounding and prediction. Analytical methods for curve fitting and trend prediction using traditional forecasting models with poor results in accuracy. In the data experiment, the evolution model automatically generated by the genetic algorithm evolutionary computer programming method was used to perform curve fitting and development trend prediction on some real historical data, and conduct in-depth analysis of feedforward and feedback errors. The results show that the evolutionary model established using this method is much more accurate than the data predicted by the three fixed traditional mathematical models of linear regression, exponential regression, and parabolic regression. Moreover, the feedforward standard deviation of the fitting curve and the predicted feedback The standard deviation is also significantly smaller.
Expand
