Processing Method And Study Of Imbalanced Data Based On Phellinus Fermentation Experiment

With the rapid development of information industry,data has become the production factor of all walks of life.Mass data is growing rapidly along with the production and operation of enterprises.A large number of new experimental data have also appeared in biological experiments.These data are characterized by high latitude,high similarity and imbalanced data.Traditional methods are difficult to cope with these problems.In this paper,we use the data collected from Phellinus fermentation experiments.And we propose an imbalanced data processing method.Finally,a classification and prediction model is established.The BP(Backpropagation Neural Network)neural network is an efficient and simple prediction model.Because the data gradient in this article is smaller,we set up a LM-BP neural network model and compared with the traditional regression equation.Genetic Algorithm is to simulate natural selection processes and genetic genetics ideas in nature by computer.Finally,we use genetic algorithm and BP neural network to optimization the Flavonoid yield.The experimental results show that the proposed based on imbalanced data processing method can be effectively used for data processing and analysis of Phellinus fermentation.The simulation of BP neural network and genetic algorithm yield prediction algorithm can be accuracy because of the error control in less than 80%.It is the basis for the design of biological experiment and the production forecast.In addition,the model proposed in this paper can optimize the yield to a certain extent,and the optimization of the output is better than the real data.