發(fā)布時間：2018-02-28 02:16

  本文關(guān)鍵詞： 臨床路徑 過程挖掘 重名活動 活動關(guān)系 統(tǒng)計α算法　出處：《浙江大學(xué)學(xué)報(工學(xué)版)》2017年10期 　論文類型：期刊論文

【摘要】：針對臨床路徑事件日志中存在的重名活動和噪音數(shù)據(jù),提出集成重名活動判別的過程挖掘算法:統(tǒng)計α算法.給出一套完整的重名活動的判別規(guī)則,用于識別過程挖掘中的重名活動并進行相應(yīng)預(yù)處理,有效地提高了過程挖掘的準(zhǔn)確性;提出基于經(jīng)典α算法改進的統(tǒng)計α算法,用于消除事件日志中各種噪音的影響.該算法在臨床路徑數(shù)據(jù)量較大的情形下,保證了結(jié)果準(zhǔn)確率和運算效率.統(tǒng)計α算法在三甲醫(yī)院的臨床數(shù)據(jù)上得到成功應(yīng)用,與經(jīng)典α算法和遺傳算法相比,該算法在效率和準(zhǔn)確性上更具優(yōu)越性.
[Abstract]:In view of the data of duplicate name activity and noise in clinical path event log, a process mining algorithm of integrating duplicate name activity discrimination is proposed: statistical 偽 algorithm, and a set of complete discriminant rules for duplicate name activity are given. It can be used to identify and preprocess the duplicate name activities in process mining, which can effectively improve the accuracy of process mining, and bring forward an improved statistical 偽 algorithm based on classical 偽 algorithm. It is used to eliminate the influence of various noises in the event log. The algorithm ensures the accuracy and efficiency of the results under the condition of large amount of data of the clinical path. The statistical 偽 algorithm has been successfully applied in the clinical data of the third Class A Hospital. Compared with classical 偽 algorithm and genetic algorithm, this algorithm is more efficient and accurate.
【作者單位】： 同濟大學(xué)機械與能源工程學(xué)院;
【基金】：國家自然科學(xué)基金資助項目(51375290) 上海航天科技創(chuàng)新基金資助項目(SAST2015054) 中央高校基本科研業(yè)務(wù)費人才資助項目
【分類號】：O213;R197.3
，

本文編號：1545326