Abstract
Exclusion tests are a well known tool in the area of interval analysis for finding the zeros of a function over a compact domain. Recently, K. Georg developed linear programming exclusion tests based on Taylor expansions. In this paper, we modify his approach by choosing another objective function and using nonlinear constraints to make the new algorithm converges faster than the algorithm in [K. Georg, A new exclusion test, J. Comput. Appl. Math. 152 (2003) 147-160]. In this way, we reduce the number of subinterval in each level. The computational complexity for the new tests are investigated. Also, numerical results and comparisons will be presented.
Original language | English |
---|---|
Pages (from-to) | 1104-1116 |
Number of pages | 13 |
Journal | Applied Mathematics and Computation |
Volume | 170 |
Issue number | 2 |
DOIs | |
Publication status | Published - Nov 15 2005 |
Keywords
- Complexity
- Exclusion tests
- Solutions of nonlinear equations
ASJC Scopus subject areas
- Computational Mathematics
- Applied Mathematics