Computing variance for interval data is NP-hard
ACM SIGACT News
Journal of Computational and Applied Mathematics - Special issue: Scientific computing, computer arithmetic, and validated numerics (SCAN 2004)
Information Sciences: an International Journal
Gradual Numbers and Their Application to Fuzzy Interval Analysis
IEEE Transactions on Fuzzy Systems
Hi-index | 0.21 |
We propose an algorithm for the computation of the interval and fuzzy variance. In particular, based on the application of the max-min inequality, we obtain an upper bound to the maximization problem and we indicate how to obtain a feasible (heuristic) solution from the upper bound. The procedure requires the minimization of a unidimensional (continuous) convex function over a compact interval, for which efficient procedures exist. Having determined the upper bound, we are able to estimate the quality of any heuristic solution. Some computational results for problems of different dimensions are reported.