TY - JOUR
T1 - Data envelopment analysis with imprecise data
T2 - An application of TAIWAN machinery firms
AU - Kao, Chiang
AU - Liu, Shiang Tai
N1 - Funding Information:
This research is supported by the National Science China under contract number NSC87-2416-H006-022.
PY - 2005/4
Y1 - 2005/4
N2 - Technology and management are two broad categories of factors that have major influence on the productivity of manufacturing firms. The ratio of the actual productivity of a firm to its maximal productivity is a measure of efficiency. Since the automation technology level and production management achievement of a firm cannot be measured precisely, a fuzzy data envelopment analysis (FDEA) model is applied to calculate the efficiency scores of a sample of machinery firms in Taiwan. Different from the conventional crisp DEA, the efficiency scores calculated from the FDEA model are fuzzy numbers, in that ranges of values at different possibility levels rather than a single crisp value are provided. When the observations are fuzzy in nature, the FDEA methodology avoids the top management from being over-confident with the results as opposed to that of using the conventional DEA methodology by simplifying the fuzzy observations to crisp values.
AB - Technology and management are two broad categories of factors that have major influence on the productivity of manufacturing firms. The ratio of the actual productivity of a firm to its maximal productivity is a measure of efficiency. Since the automation technology level and production management achievement of a firm cannot be measured precisely, a fuzzy data envelopment analysis (FDEA) model is applied to calculate the efficiency scores of a sample of machinery firms in Taiwan. Different from the conventional crisp DEA, the efficiency scores calculated from the FDEA model are fuzzy numbers, in that ranges of values at different possibility levels rather than a single crisp value are provided. When the observations are fuzzy in nature, the FDEA methodology avoids the top management from being over-confident with the results as opposed to that of using the conventional DEA methodology by simplifying the fuzzy observations to crisp values.
UR - http://www.scopus.com/inward/record.url?scp=18844415106&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=18844415106&partnerID=8YFLogxK
U2 - 10.1142/S0218488505003412
DO - 10.1142/S0218488505003412
M3 - Article
AN - SCOPUS:18844415106
SN - 0218-4885
VL - 13
SP - 225
EP - 240
JO - International Journal of Uncertainty, Fuzziness and Knowlege-Based Systems
JF - International Journal of Uncertainty, Fuzziness and Knowlege-Based Systems
IS - 2
ER -