Method and equipment for constructing classification model based on convolutional neural network

Abstract

The utility model discloses a method and equipment for constructing a classification model based on a convolutional neural network. The method comprises a step of convolution. A first stage of training is carried out on a training sample in a random convolution mode so as to acquire a convolution template value for a convolution operation, thereby constructing the classification model comprising the convolution template value. The first stage of training carried out on the training sample in the random convolution mode further comprises that connection between elements in a characteristic pattern of the current convolution layer and elements in a characteristic pattern of an upper layer which is adjacent to the current convolution layer is interrupted in a random mode based on a predetermined probability threshold value as for at least one current convolution layer. According to the invention, the number of weights used when the sample is trained can be reduced, and an over-fitting problem is relieved, thereby improving a generalization ability of the convolutional neural network.

Description

technical field [0001] The present disclosure relates to classification model construction, and more specifically, to a convolutional neural network (CNN)-based classification model construction method and device. Background technique [0002] Convolutional neural network (CNN) is a kind of artificial neural network, which has become a research hotspot in the field of speech analysis and image recognition. [0003] The convolutional layer of the traditional convolutional neural network usually has hundreds of thousands to millions of weights to learn, and at the same time, the gradient descent algorithm used in training has the problem of "gradient disappearance" in the process of backward propagation (that is, the more forward The smaller the adjustment of the gradient of one layer), this leads to a large number of weights that are not fully learned. In order to achieve a better training effect, the traditional convolutional neural network usually requires a large number o...

AI Technical Summary

Problems solved by technology

However, when the number of training samples is limited, the traditional convolutional neural network is prone to overfitting problems, resulting in poor generalization ability of the trained model.

[0004] At the same time, the sampling layer in the traditional convolutional neural network only extracts the maximum value from each local area on each feature map, which reduces the amount of calculation, but leads to excessive loss of information, because the small value in the local area It can also reflect some characteristics of this area

Images