The invention discloses a pPrivacy information protection method based on k-means clustering

Abstract

The invention belongs to the field of machine learning. The invention, and relates to a privacy information protection method based on k-means clustering. A linear homomorphic encryption algorithm LHE is combined with an addition homomorphic encryption algorithm Paillier to encrypt data; T; the method comprises the steps of obtaining ciphertext data through a cloud server, performing k-means clustering on the ciphertext data by utilizing a computing service provided by the cloud server to obtain a ciphertext clustering result, and decrypting the ciphertext clustering result by a client to obtain a plaintext clustering result. According to the invention, t, the cloud server does not obtain any privacy information of the user; t; the privacy information security of the user is ensured whilethe clustering algorithm is realized; I; in the data analysis process, data information is not leaked, t, the safety of user data is effectively improved, t, the communication traffic between the client side and the cloud server side is greatly reduced, t, the communication cost is reduced, t, the machine learning efficiency is improved, and the method is more suitable for being applied to an actual scene.

Description

technical field [0001] The invention belongs to the field of machine learning and relates to a privacy information protection method based on k-means clustering. Background technique [0002] With the rapid development of Internet technology and the resurgence of AI technology, machine learning has been widely used in the information industry, such as medical diagnosis, search engine, computer vision, detection of credit card fraud, securities market analysis, etc. The basic idea of ​​machine learning is to simulate human learning behavior, through the analysis and learning of large amounts of data, to acquire new knowledge or skills to improve the performance of existing organizational structures. More accurate machine learning results require a larger database as the learning object. However, a large amount of data contains users' private information, which brings new challenges and opportunities to the development of machine learning. Therefore, realizing data security c...

AI Technical Summary

Problems solved by technology

[0006] Usually, in order to prevent the leakage of private information, users will encrypt the data before uploading, and then upload the ciphertext to the cloud server for machine learning. The operation of machine learning brings a certain degree of difficulty

The fully homomorphic encryption algorithm supports ciphertext operations, but due to the low efficiency of fully homomorphic, this encryption method is not practical in actual scenarios

Images