Multi-dimensional Data Correlation Analysis Method Based on Neighborhood Preserving Embedding Mechanism
Due to the complexity of network structure and the existence of weak nodes, the possibility of abnormal or attacked in the network is greatly increased. Therefore, it is necessary to carry out feature extraction on the multi-dimensional data of network nodes and extract the key features, to improve...
Saved in:
Published in | 2021 IEEE International Symposium on Broadband Multimedia Systems and Broadcasting (BMSB) pp. 1 - 6 |
---|---|
Main Authors | , , , , , |
Format | Conference Proceeding |
Language | English |
Published |
IEEE
04.08.2021
|
Subjects | |
Online Access | Get full text |
Cover
Loading…
Summary: | Due to the complexity of network structure and the existence of weak nodes, the possibility of abnormal or attacked in the network is greatly increased. Therefore, it is necessary to carry out feature extraction on the multi-dimensional data of network nodes and extract the key features, to improve the speed of data analysis and find the abnormal behaviors in the network in time. In the past feature extraction methods, data fusion is often used to compress and merge multidimensional data to form new data. However, this approach cannot guarantee the influence of the original attribute dimension in the whole data set, and each dimension cannot be clearly analyzed one by one. Therefore, this paper proposes a feature extraction method based on multi-core canonical correlation analysis. The improved multi-core canonical correlation analysis algorithm is introduced into the neighborhood preserving embedding algorithm to obtain the MultiCCA-NPE feature extraction algorithm. According to the characteristics of the network operating data, in order to better extract the features, this paper first uses the improved multi-core learning CCA algorithm to calculate the correlation between the attributes of each dimension. Taking the correlation size as the basis of the NPE weight value, the improved MultiCCA-NPE algorithm is obtained, and then effective feature extraction is realized. Compared with the KCCA algorithm and the NPE-GNN algorithm, this method has a better feature extraction effect. |
---|---|
ISSN: | 2155-5052 |
DOI: | 10.1109/BMSB53066.2021.9547142 |