Complex network community discovery method based on consensus embedding

Abstract

The invention discloses a complex network community discovery method based on consensus embedding, and relates to a multi-objective optimization technology. The method comprises the following steps: 1) giving a maximum algebra maxgen and a particle swarm scale pop; 2) performing network representation learning according to a given network G = (V, E) and a network scale of n; 3) utilizing a networkto represent a learning result, initializing a particle swarm to obtain 100 particles POP, and enabling the number of iterations t to be equal to 1; 4) performing updating and variation based on consensus embedding on the POP; and 5) stopping condition: if t is less than or equal to maxgen, t<- t + 1, and turning to the step 3), otherwise, stopping and returning the Pareto frontier solution, i.e., a plurality of community division results. The updating process is more efficient and accurate, and the obtained Pareto frontier effect is more competitive; the community discovery accuracy is improved, the convergence time of the method is effectively reduced, and the method has good practicability in practical applications such as function prediction and recommendation systems.

Description

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

AI Technical Summary

Benefits of technology

This technology proposs an improved way for finding important areas within large datasets from massive amounts of data without overwhelming computational resources or requiring human input. It achieves this through a combination of algorithms embedded into a special framework called Consolidated Network Group (CN) which allows multiple agents working together efficiently while maintaining their original structure. Overall, it provides better efficiency and precision than previous methods like MODPS.

Problems solved by technology

This patented technical problem addressed in this patents relates to finding structures within complicated networks that connect similarities without being overlooked during exploration. Existing techniques like merge/split, optimizing algorithms, and simulated annealing require significant effort from both experts involved and data sources. Therefore, new ways were developed to reduce complexity while still maintaining efficiency.

Images