Skip to Main Content
 

Global Search Box

 
 
 
 

ETD Abstract Container

Abstract Header

EXPERIMENTS IN PIECEWISE APPROXIMATION OF CLASS BOUNDARY USING SUPPORT VECTOR MACHINES

KAMEI, RINAKO

Abstract Details

2003, MS, University of Cincinnati, Engineering : Computer Science.
This work is concerned with issues that arise in implementation and use of Support Vector Machines (SVM). First, an analytical computational approach to solve the convex optimization problem for SVM without any conventional heuristic optimization techniques is considered. The second issue concerns the application of SVM with linear kernels to data sets which are not linearly separable by approximating the separating surface with a collection of hyperplanes derived on small subsets of the training data. The simulation experiments on two-dimensional non-linear datasets result in good approximations of the true separating surface by the algorithm proposed. Finally, the non-linear transformation to obtain a non-linear separating surface by SVM is proposed and the piecewise linear approximation is extended to the piecewise non-linear approximation. The experimental results from both methods are compared.
Dr. Anca Ralescu (Advisor)
73 p.

Recommended Citations

Citations

  • KAMEI, R. (2003). EXPERIMENTS IN PIECEWISE APPROXIMATION OF CLASS BOUNDARY USING SUPPORT VECTOR MACHINES [Master's thesis, University of Cincinnati]. OhioLINK Electronic Theses and Dissertations Center. http://rave.ohiolink.edu/etdc/view?acc_num=ucin1061211068

    APA Style (7th edition)

  • KAMEI, RINAKO. EXPERIMENTS IN PIECEWISE APPROXIMATION OF CLASS BOUNDARY USING SUPPORT VECTOR MACHINES. 2003. University of Cincinnati, Master's thesis. OhioLINK Electronic Theses and Dissertations Center, http://rave.ohiolink.edu/etdc/view?acc_num=ucin1061211068.

    MLA Style (8th edition)

  • KAMEI, RINAKO. "EXPERIMENTS IN PIECEWISE APPROXIMATION OF CLASS BOUNDARY USING SUPPORT VECTOR MACHINES." Master's thesis, University of Cincinnati, 2003. http://rave.ohiolink.edu/etdc/view?acc_num=ucin1061211068

    Chicago Manual of Style (17th edition)