Su X Y, Li Wan Po A, Yoshioka S
School of Pharmacy, Queen's University of Belfast, Northern Ireland, United Kingdom.
Pharm Res. 1994 Oct;11(10):1462-6. doi: 10.1023/a:1018956209236.
The use of a bayesian method for estimating the shelf-life of pharmaceutical formulations is evaluated and compared with classical linear regression. Using three real data sets, the greater flexibility provided by the bayesian approach is demonstrated. In particular, the bayesian approach enabled the consideration of cases when the error distribution is non-normal. However much more computation is required with the bayesian method.
评估了一种用于估计药物制剂保质期的贝叶斯方法，并将其与经典线性回归进行比较。使用三个真实数据集，证明了贝叶斯方法具有更大的灵活性。特别是，贝叶斯方法能够考虑误差分布非正态的情况。然而，贝叶斯方法需要更多的计算。
