我用One-variable-at-a-time approach,Plackett-Burman design (PB design),Steepest ascent method,Response surface methodology (RSM)进行培养基优化，得出一个二次多项式，根据模型得出的最佳培养配比进行产量验证后，结果与模型预测相差不大。审稿人提出如下问题：Was forward and/or backward linear regression performed to evaluate the model?请问高手这是什么意思呀，我该怎么回答呢？急切盼望您的解答，谢谢！！！
Forward - add predictors one-at-a-time until no improvement in fit (Adj. R2, AIC, etc.)
Backward - start with all predictors and remove one-at-a-time until fit is hurt.
根据你的模型实际参数来回复以上两点！
回复此楼
+1/182
