基于重叠窗的稀疏评价数据计算方法

METHOD OF SPARSE EVALUATION DATA COMPUTATION BASED ON OVERLAP WINDOW

数据样本稀疏影响计算结果的准确性。针对评价数据中经常存在的数据稀疏问题,在前期研究的基础上,对原有算法进行了扩展,提出了一种在不改变原有数据集的前提下,借助窗口重叠,弥补数据稀疏,并最终提高评价计算准确度的方法。仿真实验证明,无论数据稀疏与否,重叠窗技术均可提高评价计算的准确度。简单介绍了方法的一个具体应用。

Sparsity in data sample will affect the accuracy of computation outcomes.In light of the issue that it frequently happens the phenomenon of sparsity in evaluation data,based on earlier studies,the previous method is extended and a new method is proposed to finally raise the accuracy of evaluation computation in the premise of not altering original data set and make use of overlap window technology to make up sparse data.Simulation experiment proves that the overlap window technology is effective in improving the accuracy of evaluation computation no matter whether the data is sparse or not.An actual application is introduced in brief as well.