Multi-party projection method and multi-party production data analysis method based on data security
A security and projection technology, applied in the field of data processing, can solve the problems of limited capacity, non-compliance, difficult to use, etc., and achieve the effect of solving overlapping projections, good projection effect and high security.
Active Publication Date: 2022-05-20
CENT SOUTH UNIV
View PDF4 Cites 0 Cited by
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
The multi-party t-SNE (t-Distributed Stochastic Neighbor Embedding, t-distributed random neighbor embedding) joint projection method SMAP (Secure Multi-party Projection, secure multi-party projection) based on the homomorphic encryption method can calculate the joint projection; however, the increased computational overhead of homomorphic encryption makes it difficult to apply this method in practice
The MSDSNE (Multishot Decentralized Data Stochastic Neighbor Embedding, Multishot Decentralized Data Stochastic Neighbor Embedding) projection method performs joint t-SNE projection between each data cube based on the shared anchor point data; however, there are many limitations in the MSDSNE method: First, the method needs to share The additional data set does not meet the requirements of the problem; second, the projection effect of this method is highly random; finally, the method uses the additional data set as an anchor to approximately maintain the inter-party data proximity relationship, and its ability to maintain is limited by the additional data set size
Therefore, none of the current methods can effectively solve the problem of secure multi-party projection.
Method used
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View moreImage
Smart Image Click on the blue labels to locate them in the text.
Smart ImageViewing Examples
Examples
Experimental program
Comparison scheme
Effect test
Embodiment Construction
[0056] Such as figure 1 Shown is a schematic flow chart of the method of the present invention: the multi-party projection method based on data security provided by the present invention includes the following steps:
[0057] S1. Obtain server and client sets;
[0058] S2. The server constructs the global model and the initial global dictionary, and sends the global model and the global dictionary to each client; the specific steps are as follows:
[0059] A. The server selects the model architecture and generates model parameters, builds a global model, and sends the global model to each client;
[0060] B. Each client uses the received global model to project its own local data to obtain its own local projection results;
[0061] C. Each client randomly selects a part of its local projection results and uploads them to the server;
[0062] D. The server builds an initial global dictionary based on the received projection results;
[0063] S3. Each client initializes its...
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More PUM
Login to View More Abstract
The invention discloses a multi-party projection method based on data security, which includes obtaining a set of servers and clients; the server builds a global model and an initial global dictionary and sends it to the client; The local model and the local data are projected, and the local model and some projection results are uploaded to the server; the server aggregates to obtain a new global model and a new global dictionary and sends it to the client; repeat the above steps, and the server obtains the final projection model; The service sends the projection model to the client; the client uses the projection model to project local data and uploads the results to the server; the server draws all the projection results received on a scatter diagram to complete multi-party projection. The present invention also provides a multi-party production data analysis method including the multi-party projection method based on data security. The method of the invention has good projection effect, high safety and high efficiency.
Description
technical field [0001] The invention belongs to the field of data processing, and in particular relates to a data security-based multi-party projection method and a multi-party production data analysis method. Background technique [0002] With the development of technology and the improvement of people's living standards, intelligent big data technology has been widely used in people's production and life. The current data are generally high-dimensional data, so the processing of high-dimensional data is particularly important. [0003] High-dimensional data projection method is a commonly used data analysis method. It projects high-dimensional data into a low-dimensional space, thereby supporting data analysts to analyze high-dimensional data features from low-dimensional projection results. In the past, data analysts usually collected data from multiple data providers on one device and then projected it. However, with the strengthening of people's awareness of privacy ...
Claims
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More Application Information
Patent Timeline
Login to View More Patent Type & Authority Patents(China)
IPC IPC(8): G06F16/958G06F21/62G06K9/62
Inventor 夏佳志林伟星
Owner CENT SOUTH UNIV