An Efficient LDE-Based Document Classification Algorithm
Ziqiang Wang, Xia Sun · 2009
To efficiently cope with document classification problem, an efficient document classification algorithm based on local discriminant embedding (LDE) and SVM classifier is proposed in this paper. The high-dimensional document space are first projected into the lower-dimensional feature space by using LDE algorithm, the SVM classifier is then applied in the reduced document feature space. Extensive experiments show that the proposed algorithm achieves much better performance than other traditional document classification algorithms.