Bayes框架下DNA证据的量化

Interpreting DNA Mixtures with Bayes Framework

针对DNA(脱氧核糖核酸)证据的量化过程中常用的插入算法存在量化结果与样本大小无关,而小样本时会过分量化DNA证据的缺陷,考虑样本大小的影响,引入Bayes模型,给出了基于Bayes模型下的似然比的计算公式。结合实际案例,对比了两种方法下的计算结果。结果表明基于Bayes模型下的算法比常用的插入算法更加精确合理。

Common plug-in approach neglected sample size in quantifying DNA evidence. It leaded to inaccurate results and overweight DNA evidence. The formula of calculating likelihood ratio （LR） was obtained when Bayes model was used. Two numerical examples were separately calculated according to the Bayes method and the common plug-in approach. The results showed that the Bayesian method was more conservative and accurate than that of the plug-in approach.