GBNN-填充缺失属性值算法

Based on Grey-based and kNN algorithm for imputing missing attribute values

在数据挖掘和机器学习领域,缺失数据经常出现,本文结合灰色系统理论和最近邻理论,提出了一种新的缺失数据填充方法(简称为GBNN算法),在实验中对本算法和常见的最近邻算法从分类准确率和预测正确率两个方面进行了比较,分析了本算法的优越性。

Missing data and inconsistent data has been a pervasive problem in data mining and machine learning. In this paper, we present an algorithm named GBNN（Gray-Based and kNN）, which are combined with the theory of gray-based and kNN algorithm. Experiments show that the performance of GBNN are beyond the popular algorithms, such as kNN, in prediction accuracy and classification accuracy for all kinds of real dataset in UCI.