Ako nakupovať
Doručenie
Nákupný poradca
Sparse representation of High Dimensional Data for Classification
Angličtina
137 b
30 dní na vrátenie tovaru
Mohlo by vás tiež zaujímať
In this book you will find the use of sparse §Principal Component Analysis (PCA) for representing §high dimensional data for classification. Sparse §transformation reduces the data §volume/dimensionality without loss of critical §information, so that it can be processed efficiently §and assimilated by a human. We obtained sparse §representation of high dimensional dataset using §Sparse Principal Component Analysis (SPCA) and §Direct formulation of Sparse Principal Component §Analysis (DSPCA). Later we performed classification §using k Nearest Neighbor (kNN) Method and compared §its result with regular PCA. The experiments were §performed on hyperspectral data and various datasets §obtained from University of California, Irvine (UCI) §machine learning dataset repository. The results §suggest that sparse data representation is desirable §because sparse representation enhances §interpretation. It also improves classification §performance with certain number of features and in §most of the cases classification performance is §similar to regular PCA.
Informácie o knihe
Angličtina
