不确定数据的朴素贝叶斯分类

Nave Bayes Classification for Uncertain Data

传统的机器学习和数据挖掘分类算法是在假设数据是完整精确的前提下进行的,然而在实际的应用中,由于数据存在不确定性,使这种假设很难成立.数据的不确定性可能是由多种原因导致的,比如测量错误、隐私保护以及传感器搜集的不确定信息等等.本文研究在不确定数据中使用朴素贝叶斯分类方法进行分类问题.

Traditional machine learning and data mining classification algorithms assume that data are exact and complete. However, this assumption may not hold in many situations because of data uncertainty. The uncertainty of data is arisen from measurement errors, privacy protection and information collected by sensors. In this paper, we discuss the problem of na＇fve Bayes classification for uncertain data.