Downloads

1.	One of my works is in the development of an improved algorithm for the numerical solution of the entire regularization path of the Support Vector Machine (SVM)－SVMP. The SVMP algorithm is able to handle datasets having duplicate points, nearly duplicate points and linearly dependent points, a common feature among many real world datasets. The Matlab source code for the SVMP algorithm can be downloaded here (25K Zipped). The details of the improvement can be found in our paper CJ Ong, S.Y. Shao and J.B.Yang, An improved algorithm for the solution of the entire regularization path of Support Vector Machine.
2.	One of my works is in the modification of the Gilbert-Johnson-Keerthi (GJK) algorithm for distance computation for convex objects. This resulted in the Revised GJK (RGJK) algorithm that, among various other improvements, have the advantage of almost constant time performance for incremental motions of objects. The Fortran source code for the RGJK Algorithm can be downloaded here (76K). The details of the improvement can be found in the paper
	CJ Ong and EG Gilbert, Fast Versions of the Gilbert-Johnson-Keerthi Distance Algorithm: Additional Results and Comparisons, in IEEE Transaction on Robotics and Automation, Vol. 17, No. 4, August 2001.
3.	One of my other works is in the characterization of the growth distance between two convex objects. Growth distance is a measure of both separation and penetration between two convex objects. A C version of the source code for computing the growth distance can be downloaded here (36K Zipped). The implementation of the code is described in the paper
	CJ Ong, E Huang and SM Hong, A Fast Growth Distance Algorithm for Incremental Motions, IEEE Transactions on Robotics and Automation, Vol. 16, No.6, 2000.