Differential privacy protection clustering method based on K-means

Abstract

The invention discloses a differential privacy protection clustering method based on Kmeans. The method comprises the steps of data preprocessing, clustering iteration and differential privacy; wherein each iteration needs to be carried out as follows: setting a privacy budget, calculating noise, calculating the sum of data points of each cluster and the number of points, and then adding the calculated noise to the data points; for each iteration, maintaining the upper bound and the lower bound of one privacy budget, selecting adaptive privacy budget allocation between the upper bound and thelower bound according to the availability and the privacy protection degree, and repeating the steps until the sum of squares of errors converges or reaches the upper limit of the number of iterations. According to the method, in the iteration process of the Kmeans clustering algorithm, through adaptive privacy budget allocation, data privacy is protected on the basis of data distortion, and dataavailability is guaranteed.

Description

technical field [0001] The invention belongs to the technical field of information security and relates to a clustering method and privacy protection technology, in particular to a differential privacy protection clustering method based on K-means. Background technique [0002] With the continuous popularization and deepening of information technology applications, various information systems have stored and accumulated rich data. In the face of massive data, in the industry, data holders can obtain potential value through data mining technology; in academic In the world, data mining technology has also made great progress in some research and application. As the most commonly used data mining technology, clustering algorithm is widely used. At the same time, these data contain a lot of sensitive information, which will bring immeasurable threats and losses to users. Therefore, it is necessary to protect data privacy in the process of cluster analysis. [0003] From the pe...

AI Technical Summary

Problems solved by technology

[0003] From the perspective of privacy protection, the following methods may generally be adopted: 1) Cryptography (Cryptography): the information is completely fidelity, but the computational complexity is too high

2) Anonymization: The level of privacy protection is relatively low, and may encounter NP-hard problems

The second flaw is that these early privacy-preserving models do not provide an efficient and rigorous method to demonstrate their privacy-preserving level, so when the model parameters are changed, the privacy-preserving level cannot be quantitatively analyzed

However, the traditional differential privacy algorithm is because the k-means is extremely sensitive to the initial center point, and the random noise added in the iterative process slows down the convergence speed.

Images