Community discovery method based on nonlinear non-negative matrix factorization
A technology of non-negative matrix decomposition and community discovery, applied in the field of community discovery based on nonlinear non-negative matrix decomposition, it can solve problems such as lack of learning ability and inability to improve community discovery performance, achieving the effect of strong expressive ability and improved performance
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
Method used
Image
Examples
Example Embodiment
[0042] Example 1
[0043] like figure 1 and figure 2 As shown, a community discovery method based on nonlinear non-negative matrix decomposition, including the following steps:
[0044] S1: Formation of social networks, generating no map;
[0045] S2: The root is based on the non-shadd map, and the gcn and NMF model build a unified loss function;
[0046] S3: Optimize a unified loss function to get local optimal solution;
[0047] S4: The results of the community found according to the local best solution.
[0048] In the above scheme, the introduction of the map volume network GCN and the NMF model build a unified loss function, and combined training to the NMF nonlinear characteristics, thereby improving the improvement of NMF community discovery method performance and strong expression.
[0049] Preferably, in step S1, the non-paragraph G = (V, E), where V = {V 1 , ..., v n } Collection of n network user nodes, e = {e ij| v i ∈V∧V j ∈V} is represented as a collection of connect...
Example Embodiment
[0070] Example 2
[0071] figure 2 It is the overall model frame of the method comprising an input layer, a map volume layer, and an output layer, wherein the input layer employs the adjacent matrix and the unit matrix of the graph, and the map volume layer uses a 2 layer diagram convolution network, the activation function is RELU The output layer is responsible for outputting the map node that is normalized after the SOFTMAX function represents Z and is used to rebuild the adjacency matrix. The entire model framework is trained through the reconstruction of the map and the NMF combined loss function.
Example Embodiment
[0072] Example
PUM
Abstract
Description
Claims
Application Information
- R&D Engineer
- R&D Manager
- IP Professional
- Industry Leading Data Capabilities
- Powerful AI technology
- Patent DNA Extraction
Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic.
© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap