Robust scale estimators for fuzzy data

摘要: Observations distant from the majority or deviating general pattern often appear in datasets. Classical estimates such as sample mean variance can be substantially affected by these observations (outliers). Even a single outlier have huge distorting influence. However, when one deals with real-valued data there exist robust measures/estimates of location and scale (dispersion) which reduce influence atypical values provide approximately same results classical applied to typical without outliers. In real-life, analyzed interpreted are not always precisely defined they cannot properly expressed using numerical measurement. Frequently, some imprecise could suitably described modelled considering fuzzy rating this paper, several well-known estimators case extended for random numbers (i.e., mechanisms generating fuzzy-valued data), their properties dispersion examined. Furthermore, behaviour is two powerful tools, namely, finite breakdown point sensitivity curves. Simulations, including empirical bias curves, performed complete study.