Novel transductive semi-supervised data classification method and system

Abstract

The invention discloses a novel transductive semi-supervised data classification method and system, which integrate the unsupervised subspace feature learning, discriminant clustering and adaptive semi-supervised classification into a unified framework seamlessly, and perform semi-supervised learning based on low-dimensional manifold features of original data and discriminant subspace clustering results, and can be used for high-dimensional data representation and classification. Based on the above-mentioned joint model, the graph construction and label propagation process are also seamlesslycombined, and thus, an adaptive weight coefficient matrix based on the low-dimensional manifold features and soft category labels of unlabeled data can be obtained.

Description

technical field [0001] The invention relates to a new transductive semi-supervised data classification method and system, belonging to the technical fields of data mining and computer vision. Background technique [0002] With the continuous development of computer technology and intelligence, most of the real data generated in our daily life and communication are usually not easy to distinguish due to the lack of identification information (such as class information). In addition, the labeling process of data is also expensive and time-consuming, and applying fully supervised methods to obtain all data labels requires a large overhead. As a result, semi-supervised learning methods that can use both a small amount of labeled data and a large amount of unlabeled data have attracted increasing attention in recent years. Therefore, how to effectively use a small amount of labeled information to improve classification accuracy is a problem that needs to be further explored. ...

AI Technical Summary

Problems solved by technology

[0005] The purpose of the present invention is to provide a new transductive semi-supervised data classification method and system to overcome the problem of high overhead in obtaining data labels in the prior art

Images