The Research of Nonlinear Dimensionality Reduction with Isomap and C-Isomap
YUAN Li-guo1,TANG Wu-lei2(1.Department of Mathematics,South China Agricultural University,Guangzhou 510642 China;2.Gudong Soft Park,Guangzhou 510663 China)
The paper mostly introduces the thinking, advantage and disadvantage of the Isomap algorithm. Clustering algorithm was applied to cluster the sampled data and Then kernel function was utilized to adjust the Euclidean distance between data points, so the Isomap algorithm based clustering can improve the performance and the range of application of Isomap. In the end of paper, the sampled data of Swiss-Roll was utilized to test the algorithm of Isomap and C-Isomap, the result has proved that C-Isomap can better discover the dimensionality reduction.