【摘要】 将最大期望值算法(EM)与朴素贝叶斯算法(NB)相结合,提出EM-NB算法来填补网络调查中的缺失数据。对比基于处理后的完备数据集的分类统计结果与基于纸质调查得到的分析结果,结果显示,利用EM-NB算法处理缺失数据后的网络调查问卷与纸质调查问卷可得到一致的调查结果。这表明EM-NB算法是一种有效的处理网络调查中缺失数据的方法。更多还原

【Abstract】 Combining the expectation maximization(EM)algorithm with the Naive Bayes(NB)algorithm,this paper proposes the EM-NB algorithm which is used to process the missing data in network investigation.Then it compares the sub-category statistical analysis result based on the treated complete data set with the analysis results based on the paper-based survey by using an example.It finds that the analysis results based on these two groups of survey data are coincident,which shows that the EM-NB algorithm is an effective method processing the missing data in network survey.更多还原