人工智能辅助下残缺数据样本集补全算法与应用  

Artificial Intelligence Assisted Incomplete Data Sample Set Completion Algorithm and Its Application

作　　者：李洋[1] 张镝[1] LI Yang;ZHANG Di(Clinical Medicine School,Changchun Medical College,Changchun 130031,China)

机构地区：[1]长春医学高等专科学校,临床医学院,吉林长春130031

出　　处：《微型电脑应用》2025年第1期58-60,64,共4页Microcomputer Applications

基　　金：吉林省教育厅科学研究项目(JJKH20231542KJ)。

摘　　要：在补全残缺数据样本集的过程中,由于缺少评价数据样本集合的相似度导致数据值估计准确率低、补全程度低等问题,提出一种新的残缺数据样本集补全算法。通过插值模型构建残缺数据拟合函数,得到相似数据样本集。通过皮尔森相关系数评价相似数据样本集的相似度,得到残缺数据样本集补全权重。采用推荐算法计算最优推荐数值,实现残缺数据样本集的补全。实验结果表明,与现有残缺数据样本集补全算法相比,所提算法极大地提升了数据值的估计准确率与补全率,充分说明该算法具备更好的补全性能,能够保证各领域数据的完整性,具有较强的实际应用性。To address the issues of low accuracy and faults in estimating data values during the process of completing incomplete data sample sets due to the lack of similarity of evaluation data sample set,a new algorithm for completing the incomplete data sample set is proposed.An interpolation model is used to construct a fitting function for the incomplete data,and obtain a set of similar data samples.The Pearson correlation coefficient is used to evaluate the similarity of the similar data sample set,and obtain weights for completing the incomplete data sample set.A recommendation algorithm is used to calculate the optimal recommended values,achieving the completion of the incomplete data sample set.Experimental results show that compared to existing algorithms for completing incomplete data sample sets,this algorithm greatly improves the accuracy of data value estimation and completion rate,which fully demonstrates its superior completion performance and strong practical application capabilities in ensuring data integrity in various fields.

关 键 词：人工智能 残缺 数据样本集 数据补全 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]