Filtering Domains of Factorable Functions Using Interval Contractors
Many theorems in mathematics require a real function to be continuous over the domain under consideration. In particular the Brouwer fixed point theorem and the mean value theorem underlie many interval methods like Newton operators for solving numerical constraint satisfaction problems or global optimization problems. Since the continuity property collapses when the function is not defined at some point it is then important to check whether the function is defined everywhere in a given domain. We introduce here an interval branch-and-contract algorithm that rigorously approximate the domain of definition of a factorable function within a box. The proposed approach mainly relies on interval contractors applied to the domain constraints and their negations stemming from the expression of the function.
KeywordsInterval methods Branch-and-contract algorithm Interval contractor Constraint satisfaction problem Paving
The author would like to thank Christophe Jermann for interesting discussions about these topics and his careful reading of a preliminary version of this paper.
- 1.1788-2015, I.S.: IEEE Standard for Interval Arithmetic (2015)Google Scholar
- 3.Benhamou, F., Goualard, F., Granvilliers, L., Puget, J.F.: Revising hull and box consistency. In: Proceedings of International Conference on Logic Programming (ICLP), pp. 230–244 (1999)Google Scholar
- 6.Collavizza, H., Delobel, F., Rueher, M.: Extending consistent domains of numeric CSP. In: Proceedings of International Joint Conference on Artificial Intelligence (IJCAI), pp. 406–413 (1999)Google Scholar
- 8.Granvilliers, L.: A new interval contractor based on optimality conditions for bound constrained global optimization. In: Proceedings of International Conference on Tools with Artificial Intelligence (ICTAI), pp. 90–97 (2018)Google Scholar
- 11.Lhomme, O.: Consistency techniques for numeric CSPs. In: Proceedings of International Joint Conference on Artificial Intelligence (IJCAI), pp. 232–238 (1993)Google Scholar
- 13.Moore, R.E.: Interval Analysis. Prentice-Hall (1966)Google Scholar