本文選題：RFID數(shù)據(jù)流　切入點：不確定性　出處：《計算機科學》2016年S2期

【摘要】：原始RFID數(shù)據(jù)流含有大量噪聲且具有不確定性,必須在使用之前對其進行數(shù)據(jù)清洗,而清洗策略是清洗質(zhì)量的保證。提出一種適合不確定RFID數(shù)據(jù)流的清洗策略。該清洗策略引入了最大熵原理,對待清洗的RFID元組的特征屬性進行權重選擇,并根據(jù)清洗節(jié)點的時間消耗以及誤差進行清洗成本分析,決策出最佳的清洗方法。仿真實驗結果表明,該清洗策略提高了不確定RFID數(shù)據(jù)流的清洗效率與精度。
[Abstract]:The raw RFID data stream contains a lot of noise and uncertainty and must be cleaned before it is used. The cleaning strategy is the guarantee of cleaning quality. A cleaning strategy suitable for uncertain RFID data flow is proposed. The maximum entropy principle is introduced in the cleaning strategy, and the weight of the characteristic attributes of the RFID tuple is selected. According to the time consumption and error of cleaning node, the best cleaning method is determined. The simulation results show that the cleaning strategy improves the cleaning efficiency and precision of uncertain RFID data stream.
【作者單位】： 南京森林警察學院信息系;南京理工大學計算機學院;
【基金】：中央高�；究蒲袠I(yè)務(LGYB201602)資助
【分類號】：TP391.44
，

本文編號：1686586