The complexity of computation and approximation of the t-ratio over one-dimensional interval data

Michal Černý and Milan Hladík

Computational Statistics & Data Analysis, 2014, vol. 80, issue C, 26-43

Abstract: The main question is how to compute the upper and lower limits of the range of possible values of a given statistic, when the data range over given intervals. Initially some well-known statistics, such as sample mean, sample variance or F-ratio, are considered in order to illustrate that in some cases the limits can be computed efficiently, while in some cases their computation is NP-hard. Subsequently, the t-ratio (variation coefficient) is considered. It is investigated when the limits for t-ratio are computable in polynomial time and a new efficient algorithm is designed for this case. Conversely, complementary NP-hardness results are proved, demonstrating the cases when the computation cannot be done efficiently. Subsequently, the NP-hardness results are strengthened: it is shown that under certain assumptions, even an approximate evaluation with an arbitrary absolute error is NP-hard. Finally, it is shown that the situation can also be (in some sense) regarded positively: a new pseudopolynomial algorithm is developed. The algorithm is of practical importance, especially when the dataset to be processed is large and does not contain “excessively” large numbers.

Keywords: Interval data; t-ratio; Computational complexity; NP-hardness; Inapproximability; Pseudopolynomial algorithms (search for similar items in EconPapers)
Date: 2014
References: View references in EconPapers View complete reference list from CitEc
Citations: View citations in EconPapers (1)

Downloads: (external link)
http://www.sciencedirect.com/science/article/pii/S0167947314001789
Full text for ScienceDirect subscribers only.

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:eee:csdana:v:80:y:2014:i:c:p:26-43

DOI: 10.1016/j.csda.2014.06.007

Access Statistics for this article

Computational Statistics & Data Analysis is currently edited by S.P. Azen

More articles in Computational Statistics & Data Analysis from Elsevier
Bibliographic data for series maintained by Catherine Liu ().