Premium
THE USE OF FORECAST ERROR MEASURES AS SURROGATES FOR AN ERROR COST CRITERION IN THE PRODUCTION SMOOTHING PROBLEM
Author(s) -
CHENTNIK CHESTER G.
Publication year - 1972
Publication title -
decision sciences
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.238
H-Index - 108
eISSN - 1540-5915
pISSN - 0011-7315
DOI - 10.1111/j.1540-5915.1972.tb00536.x
Subject(s) - smoothing , mean squared error , approximation error , statistics , bayes error rate , mathematics , forecast error , selection (genetic algorithm) , measure (data warehouse) , mean squared prediction error , mathematical optimization , computer science , econometrics , data mining , bayesian probability , bayes' theorem , bayes classifier , artificial intelligence
This paper presents research on the problem of selecting a proper surrogate for a forecast error cost criterion in the production smoothing problem. Various forecast models estimated future selected demand process values. Resultant error costs were computed and the coincidence of the selection of a forecast model on the basis of least error cost and the various error measures was noted. The error measures used were the mean absolute deviation, average algebraic error(bias), and the mean squared error. Computations necessary to develop the mathematical form of the error cost criterion are presented in an Appendix. Also presented are the penalty costs of using an error measure as a surrogate for an error cost criterion.