Fuzzy Systems and Knowledge Discovery, Fourth International Conference on (2007)
Haikou, Hainan, China
Aug. 24, 2007 to Aug. 27, 2007
DOI Bookmark: http://doi.ieeecomputersociety.org/10.1109/FSKD.2007.459
Zhenyue Zhang , Zhejiang university, China
Lingxiao Zhao , Zhejiang university, China
We propose a novel dimension reduction method for clas- sification using a probability-based distance and the tech- nique of locally linear embedding (LLE). Logistic Discrim- ination (LD) is adopted for estimating the probability dis- tribution as well as for classification for the reduced data. Different to the supervised locally linear embedding (SLLE) that is only used for the dimension reduction of train- ing data, our probability-based locally linear embedding (PLLE) can be applied on both training and testing data. Five microarray data sets in high dimensional spaces, the IRIS data, and a real set of handwritten digits are experi- mented. The numerical results show that our method per- forms better, compared with the LD classifiers applied on the LLE or SLLE mapped lower dimensional data.
L. Zhao and Z. Zhang, "Probability-Based Locally Linear Embedding for Classification," 2007 International Conference on Fuzzy Systems and Knowledge Discovery(FSKD), Haikou, 2007, pp. 243-247.