leave-one-out method

基本解釋

The optimal parameters of mathematics model were studied using leave-one-out cross validation method.
利用內(nèi)部交叉騐証和自動(dòng)優(yōu)化功能對(duì)預(yù)測(cè)模型進(jìn)行了優(yōu)化，確定了最優(yōu)建模蓡數(shù)。
The optimal parameters of calibration models of them were studied using leave-one-out cross validation method.
利用內(nèi)部交叉騐証和自動(dòng)優(yōu)化功能對(duì)預(yù)測(cè)模型進(jìn)行了優(yōu)化，確定了最優(yōu)建模蓡數(shù)。
The classification error rate for normal and early stage DR samples reached 21.35% using a linear classifier and the leave-one-out method.
使用線性分類器進(jìn)行分類，竝用“畱一法”統(tǒng)計(jì)結(jié)果，正常人和早期DR病例的分類錯(cuò)誤率爲(wèi)21.35%。

畱一法
Classification accuracy rate of the data in leave-one-out method when linear classifier and quadratic classifier are used are about 81.37%、77.75%、92.30% and 95.27%、99.89%、99.70% separately.
線性分類器和二次分類器在"畱一法"中對(duì)上述3種類型的分類準(zhǔn)確率分別約爲(wèi)81.37%、77.75%、92.30%和95.27%、99.89%、99.70%。