Computer Science
6 years ago

New method boosts accuracy and speed of pattern recognition technology.

IEEE Access
Multiple Kernel-Based Discriminant Analysis via Support Vectors for Dimension Reduction
Shan Zeng, Chongjun Gao, Xiuying Wang, Liang Jiang, Dagan Feng

Paper Summary

Related papers

Theme background image
Computer Science
8 years ago

New orthogonal methods revolutionize solving complex classification problems.

Dimensionality Reduction
Chemometrics
Hyperspectral Analysis
Theme background image
Earth and Planetary Sciences
8 years ago

New method boosts classification accuracy for complex data sets

Remote Sensing
1 view
Theme background image
Computer Science
8 years ago

New KNDA method revolutionizes data analysis, promising better results for all.

Dimensionality Reduction
Neural Networks
Image Compression

The article introduces a new method called MKDASV that combines multiple kernels to improve pattern recognition. By using support vectors, the method effectively represents data structure and reduces computational burden. Experimental results show that MKDASV outperforms other methods in terms of accuracy and efficiency.