Short communication Classification methods for multivariate quality parameters
| Autor: | W. Baumeister, H. Schmidli, C. Weihs |
|---|---|
| Rok vydání: | 1993 |
| Předmět: |
Multivariate statistics
Computer science Applied Mathematics media_common.quotation_subject Decision rule computer.software_genre Linear discriminant analysis Data structure Cross-validation Analytical Chemistry Statistical classification Data quality Quality (business) Data mining computer media_common |
| Zdroj: | Journal of Chemometrics. 7:131-142 |
| ISSN: | 1099-128X 0886-9383 |
| Popis: | There are many chemical products where product conformity is decided upon by qualitative human judgements of overall product quality. Nowadays, quantitative instrumentally determined quality parameters become available which are intended to replace such qualitative judgements by means of automatic decision rules using multivariate specification limits. Six classification methods to derive such limits are compared in terms of their power to predict corresponding human judgemets on overall color conformity of 17 dyestuffs based on historical quality data. Standard statistical classification methods turned out to be unacceptable for the routine generation of decision rules because of the frequent distinct suboptimality of their predictive power. Instead, a simple non-statistical classification method utilizing a priori knowledge about the underlying data structure yielded uniformly satisfactory decision rules. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|
Plný text