判别分析法在小麦新陈度判别中的应用

Application of discriminant analysis in distinguishing the freshness of wheat

小麦新陈度直接影响着小麦后期交易和面制品质量,通过对不同年份300多个小麦样品的研究,选择了合适判别指标,利用Bayes判别分析法建立了两类小麦新陈度判别模型,利用自身验证法和交叉验证法检验了模型的稳定性。结果表明:按照新陈小麦两类划分模型的正确率高达96%,能满足实际应用;按照实际年份划分效果不理想。

Wheat freshness ＂affects the trading of wheat and the quality of flour products directly. Two kinds of discriminant function for wheat freshness were established by the Bayes discriminant analysis, with 300 wheat samples in different storage periods, and some suitable diseriminant index. The stability of the models was checked up by self validation and cross validation. The results showed that the correct rate of the model according to wheat freshness was 96% ,which was able to use in practice. While that of the other model according to harvest years was not satisfied.