Based On The Modified GWO-BP Algorithm Model Of DNA Sequence Classification Study

The study of DNA sequence classification can help humans to explore unknown species,which is of great significance for discovering new species and making reasonable classification and prediction,and also has important research value in medical and environmental science.With the development of science and technology,the drastic expansion of diversified biological information has produced a huge amount of data.Therefore,in order to explore the law of gene composition,how to extract and mine effective information from these data and choose what method to process information has become a research problem.This paper presents a DNA sequence classification model based on Back Propagation(BP)neural network,aiming to solve the accuracy of DNA sequence classification.The model introduced the grey Wolf algorithm to optimize the initial weight and threshold of traditional BP neural network,expand the global search ability of the model,and the grey Wolf algorithm,balance the global and local search ability of the grey Wolf algorithm,make the grey Wolf algorithm more accurately to find the optimal weight and threshold,and then use the test function to improve the grey Wolf algorithm simulation test,confirmed the feasibility of improving the grey Wolf algorithm.Second,according to the statistical theory of DNA sequence feature extraction,single base frequency,double base frequency,three base frequency fusion of35-dimensional feature vector as expression DNA sequence,select principal component analysis(Principal Component Analysis,PCA)to the initial feature vector to reduce dimension,avoid redundant features affect the training accuracy of the model,finally get 12-dimensional feature vector.Build the data set and train the model,and compare the curves of error function and classification accuracy data statistics.Finally,the DNA sequence classification model of BP neural network based on the improved grey Wolf algorithm.The experimental results show that the DNA sequence classification model based on the improved grey Wolf algorithm has higher accuracy than the PSO-BP and GWO-SVM classification models.In this paper,we propose a theoretical improvement on the defects of traditional BP neural network and traditional grey Wolf algorithm,combined with the model building,and finally establish and verify the dataset.For this experiment,there are inevitably some shortcomings,where the convergence rate of the functions in the model needs to be improved,and the local extreme value is not excluded in the grey Wolf algorithm.Therefore,the improvement and innovation research of neural network technology still need to continue,but as one of the main components in the field of artificial intelligence,the prospect of neural network technology is still very considerable.