Target screening method and system based on low-rank sparse joint tensor of enhanced latent space

By projecting multi-view data into the latent space and separating the block diagonal and non-block diagonal structures of the self-representation matrix, and employing improved tensor kernel norm and sparse norm constraints, the problem of not capturing high-order and local correlations in the prior art is solved, thereby improving the accuracy of target screening.

CN122223376BActive Publication Date: 2026-07-21SOUTHWEAT UNIV OF SCI & TECH
View PDF 1 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
SOUTHWEAT UNIV OF SCI & TECH
Filing Date
2026-05-15
Publication Date
2026-07-21

AI Technical Summary

Technical Problem

Existing technologies fail to effectively capture higher-order and local correlations during target screening, resulting in low accuracy.

Method used

By projecting a multi-view data matrix into the latent space, the block diagonal and non-block diagonal structures of the self-representation matrix are separated. The objective function is constructed using tensor double arctangent kernel norm and sparse norm constraints. The alternating direction multiplier method is used for optimization, and finally, the spectral clustering algorithm is used for target selection.

Benefits of technology

It improves the accuracy of image clustering and ensures the accuracy of target selection, especially performing well in application scenarios such as ore screening and animal and plant classification.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN122223376B_ABST
    Figure CN122223376B_ABST
Patent Text Reader

Abstract

The application discloses a target screening method and system based on a low-rank sparse joint tensor of an enhanced latent space, relates to the technical field of target screening, and comprises the following steps: introducing a projection matrix to project a first matrix into a latent representation matrix, and performing separation of a noise matrix once to obtain a second matrix and a first noise matrix; performing self-representation on the second matrix in a latent space, and performing separation of a noise matrix twice to obtain a self-representation matrix of multi-view data and a second noise matrix; performing separation of a block diagonal and a non-block diagonal on the self-representation matrix, and constructing a block diagonal tensor and a non-block diagonal tensor; respectively calculating norms of the noise matrix, the block diagonal tensor and the non-block diagonal tensor, and performing weighted summation on the respective norm results to obtain a target function; solving the target function by using an alternating direction multiplier method; and obtaining a multi-view self-representation coefficient matrix when the solution result of the target function is the minimum value.
Need to check novelty before this filing date? Find Prior Art

Citation Information

Patent Citations

  • Low-dimensional subspace clustering method based on projection matrix guidance

    CN121330328A