Research And Application Of Deep Learning Algorithm Based On Cloud Model Principal Component

Depth study is the focus of the current study,In this paper,based on the previous studies,the traditional convolution neural network algorithm can not be independent study in the convolution kernel number determined,initial parameters and the channel choose between the network layer and layer,is proposed a new neural network training method based on independent component analysis and cloud model.Firstly,the relationship between the main vector of the principal component analysis and the convolution kernel number of the convolution neural network is analyzed.Then,the independent principal component neural network method is used to analyze the unlabeled image mixed signal to determine the depth of the convolution kernel,and the fast fixation algorithm Initialization of convolution kernel parameters;Finally,the cloud model selector is used to select the channel connection between the model layer and the layer,so as to reduce the number of network training steps.The experimental results show that the average principal vector approximation model of the sampled samples obtained by principal component analysis is the inflection point of the number of convoluted nuclei in the case that the accuracy of the test set is basically invariant,and the number of convolution nuclei is determined by the repeated test.The independent algorithm and the cloud model selector make the model training step reduce and accelerate the model convergence.The proposed algorithm is about 5.4% higher than the classical algorithm.The accuracy of the test set is basically the same as that of the original convolution neural network.