Supervised Kernel Locally Linear Embedding
Shen Jia-feng · Science Technology and Engineering · 2008
Kernel locally linear embedding is an excellent manifold learning methods.It perform well for non-linear data dimensionality reduction.But the methods is an unsupervised learning algorithms,its effect is not good for supervised learning problems,such as pattern recognition.By analyzing the mechanism of supervised learning,a supervised kernel locally linear embedding method is proposd.Data experiment show that the algorithm performs well for supervised learning problems.