Abstract | ||
---|---|---|
In this paper we deal with the solution of the almost Toeplitz tridiagonal systems that arise from the problem of curve fitting by Natural Cubic Splines and B-Splines. We propose the TJ decomposition that gives rise to a method which is more accurate and faster than other previously proposed methods as we prove along the work. For the solution of the recurrences that arise from the TJ decomposition, we propose a specialization of the Overlapped Partitions Method (OPM). We show that OPM compares favorably in the context of the problem to the classic Divide and Conquer and R-Cyclic Reduction on the Convex C-3480 supercomputer. |
Year | DOI | Venue |
---|---|---|
1996 | 10.1109/EMPDP.1996.500611 | PDP |
Keywords | Field | DocType |
tj decomposition,toeplitz tridiagonal system,curve fitting,natural cubic splines,convex c-3480 supercomputer,parallel interpolation,r-cyclic reduction,overlapped partitions method,biomedical imaging,cubic spline,design automation,animation,graphics,interpolation,divide and conquer,parallel algorithms,shape,spline,b splines,context modeling | Tridiagonal matrix,Spline (mathematics),Applied mathematics,Mathematical optimization,Curve fitting,Parallel algorithm,Interpolation,Regular polygon,Toeplitz matrix,Divide and conquer algorithms,Mathematics | Conference |
Citations | PageRank | References |
1 | 0.41 | 9 |
Authors | ||
3 |
Name | Order | Citations | PageRank |
---|---|---|---|
josepl larribapey | 1 | 1 | 0.41 |
Juan J. Navarro | 2 | 323 | 42.90 |
Àngel Jorba | 3 | 15 | 3.61 |