An image classification method based on differential privacy and hierarchical correlation propagation

Abstract

The invention discloses an image classification method based on differential privacy and hierarchical correlation propagation, belonging to the technical field of data security. The idea is as follows: determining a gray image data set D, wherein the gray image data set D comprises m gray image data sets; calculating the correlation matrix R of the grayscale image data set D and the noise averagecorrelation matrix R(bar) of the grayscale image data set D; setting the convolution neural network comprising num_conv convolution layers and num_FC full connection layers, wherein theta denotes allparameters of the convolution neural network, theta= {theta <Conv>, theta <FC>}, theta <Conv> denotes parameters of num_conv convolution layers of the convolution neural network, and theta <FC> denotes parameters of num_FC full connection layers of the convolution neural network; further obtaining the optimal parameter theta (hat), theta (hat)={theta (hat)<conv>, theta(hat)<FC>} of the convolutionneural network, wherein theta (hat)<conv> denotes the optimal parameters of convolution neural network num_conv convolution layers, and theta(hat)<FC> denotes the optimal parameters of num_FC full connected layers of a convolution neural network; taking the optimal parameter of convolution neural network num_conv convolution layers {theta (hat)<conv> and the optimal parameter of convolution neural network num_FC convolution layers theta(hat)<FC> as an image classification result based on differential privacy and hierarchical correlation propagation.

Description

technical field [0001] The invention belongs to the technical field of data security, in particular to an image classification method based on differential privacy and hierarchical correlation propagation. Background technique [0002] With the widespread application of artificial intelligence in recommendation systems, its efficient information filtering technology enables users to obtain the products and information they are interested in more efficiently. However, when recommending systems to users, there is a risk of violating user privacy. On the one hand, users may worry that the recommended results contain too much content, which will reveal their private information; on the other hand, users who are interested in the recommended results may worry that the recommended content contains vulgar information, which will limit their ability to improve their technology; therefore, Dealing with private data usually needs to consider the balance between data availability and p...

AI Technical Summary

Problems solved by technology

[0002] With the wide application of artificial intelligence in the recommendation system, its efficient information filtering technology enables users to obtain the products and information they are interested in more efficiently, but there is a risk of violating user privacy when the recommendation system recommends to users

On the one hand, users may worry that the recommended results contain too much content, which will reveal their private information; on the other hand, users who are interested in the recommended results may worry that the recommended content contains vulgar information, which will limit their ability to improve their technology; therefore, Dealing with private data usually needs to consider the balance between data availability and privacy protection

[0003] In the whole big data marketing, images, as an important part of user behavior data, are widely used in major e-commerce websites. The recommendation system extracts the content features of images through the model analysis module, such as low-level features such as color, texture, and shape. PHOG , SIFT, CEDD and other advanced features, and then classify and model the image, calculate the similarity with the image in the same classification library, and finally make intelligent recommendations based on the similarity degree and related matching degree of the image; however, the current image data-based The recommendation system does not protect the privacy of the data

Images