online read us now
Paper details
Number 2 - June 2020
Volume 30 - 2020
Flexible resampling for fuzzy data
Przemyslaw Grzegorzewski, Olgierd Hryniewicz, Maciej Romaniuk
Abstract
In this paper, a new methodology for simulating bootstrap samples of fuzzy numbers is proposed. Unlike the classical
bootstrap, it allows enriching a resampling scheme with values from outside the initial sample. Although a secondary
sample may contain results beyond members of the primary set, they are generated smartly so that the crucial characteristics
of the original observations remain invariant. Two methods for generating bootstrap samples preserving the representation
(i.e., the value and the ambiguity or the expected value and the width) of fuzzy numbers belonging to the primary sample
are suggested and numerically examined with respect to other approaches and various statistical properties.
Keywords
bootstrap, fuzzy data, fuzzy numbers, fuzzy sample, imprecise data, resampling