Research On Application Of Machine Learning Algorithm In Data Classification

Many practical problems in reality can be transformed into data classification in data information processing,such as weather forecast,commodity recommendation,biological information,network detection,and data processing are based on machine learning.The development of science and technology,machine learning algorithm applications can also be very broad.This paper mainly introduces two machine learning algorithms: particle swarm optimization algorithm to support vector machines and convolution neural networks.(1)Based on the characteristics of various leaves,a data preprocessing model is constructed: the first part of the paper is used to solve the problem of genetic classification.The principal component analysis method was used to extract the three principal components from the 16 features,and then the support vector machine(SVM)was optimized by particle swarm optimization.The support vector machine was used to predict the leaf data.The experimental results show that the particle swarm optimization algorithm is highly accurate and up to 94.1%,which is higher than the other two classification methods,compared with the genetic algorithm and the grid search method.(2)The particle swarm optimization model of support vector machine was applied to the classification of cancer genes.By selecting several different experimental data,the different classification effects of three different classification methods on cancer gene classification were analyzed.As a result,the particle swarm optimization support vector machine achieves the best classification effect in the three classification methods.(3)The convolution neural network is applied to the image processing,and by optimizing the convolution of the convolutional neural network and the filter in the pooling layer Function,to optimize the performance of the role,and the number of parameters to a minimum,and then construct a certain structure of the convolution neural network,and then the model of the image data set classification processing,the final image to achieve thedesired classification results.