Advances and Applications in Statistics
Volume 6, Issue 2, Pages 207 - 216
(August 2006)
|
|
SUPPORT VECTOR MACHINE WITH KERNEL METHODS AND SIMULATIONS
Tae-Soo Kim (South Korea) and Jung-Ho Ahn (South Korea)
|
Abstract: SupportVectorMachines(SVMs)arepowerfultoolsfordataclassification. SVMs attempt to separate two given sets in N-dimensional real (Euclidean) space
by a nonlinear surface, often only implicitly defined by a kernel function. We examined the priority of given various kernel functions for given data sets which follow particular probability distributions. |
Keywords and phrases: support vector machine, kernel method, Lagrangian, risk function. |
|
Number of Downloads: 359 | Number of Views: 1211 |
|