Disciplines
Other Social Sciences (25%); Computer Sciences (50%); Mathematics (25%)
Keywords
Non-Precise Data,
Application of Fuzzy Models,
Characterizing Functions,
Measurement Analysis,
Statistical Methods,
Efficient Algorithms
Final report
Measurement results of continuous quantities are always more or less non-precise. This imprecision should not be
confused with variability and errors. Imprecision (Fuzziness) is a feature of individual measurements and has to be
described quantitatively in order to avoid unrealistic data analysis results, even in case of precision measurements.
This description is possible using the concept of so-called non-precise numbers and corresponding characterizing
functions.
The projects most important result is the possibility to describe and analyze non-precise data using software which
has been implemented.
In order to make this available, firstly it was necessary to generalize statistical methods to the situation of non-
precise data. This was done at the Institute of Statistics and Probability Theory. Then the generalized analysis
methods were implemented at the institute.
Possible applications are in all scientific fields, where measurements are essential, and where imprecision cannot
be neglected, which range from the field of medicine via social and ecological disciplines to technological
analyses.