基于EMD和ELM相結(jié)合的門診量預(yù)測(cè)模型研究
網(wǎng)絡(luò)安全與數(shù)據(jù)治理 6期
樊沖
(錦州市大數(shù)據(jù)管理中心,遼寧錦州121000)
摘要： 針對(duì)門診量波動(dòng)幅度較大的時(shí)間序列預(yù)測(cè)問(wèn)題，先采用經(jīng)驗(yàn)?zāi)B(tài)分解(EMD)將非線性較強(qiáng)的原始數(shù)據(jù)進(jìn)行分解，然后通過(guò)極限學(xué)習(xí)機(jī)(ELM)將分解后的各個(gè)序列分量進(jìn)行建模，最后將各個(gè)分量的預(yù)測(cè)值相加得出最終結(jié)果。將BP神經(jīng)網(wǎng)絡(luò)、ELM兩個(gè)單一模型與EMDELM組合模型進(jìn)行對(duì)比驗(yàn)證，實(shí)驗(yàn)結(jié)果表明組合模型的精準(zhǔn)度明顯好于兩個(gè)單一模型。
中圖分類號(hào)：TP391
文獻(xiàn)標(biāo)識(shí)碼：A
DOI:10.19358/j.issn.2097-1788.2023.06.016
引用格式：樊沖.基于EMD和ELM相結(jié)合的門診量預(yù)測(cè)模型研究［J］.網(wǎng)絡(luò)安全與數(shù)據(jù)治理，2023，42（6）：97-102.
文獻(xiàn)標(biāo)識(shí)碼：A
DOI:10.19358/j.issn.2097-1788.2023.06.016
引用格式：樊沖.基于EMD和ELM相結(jié)合的門診量預(yù)測(cè)模型研究［J］.網(wǎng)絡(luò)安全與數(shù)據(jù)治理，2023，42（6）：97-102.
Research on outpatient volume prediction model based on the combination of EMD and ELM
Fan Chong
（Jinzhou Big Data Management Center，Jinzhou 121000，China）
Abstract： Aiming at the time series prediction with largefluctuations of outpatient volume, firstly, it is necessary to decompose original data with strong nonlinearity by Empirical Mode Decomposition (EMD), model these decomposed sequence components by Extreme Learning Machine (ELM), and then sum up the prediction volume of these sequence components and finally draw a conclusion. The single models of BP neural network and ELM were compared and verified with the combined model of EMDELM, and it was found that the accuracy of the combined model was significantly better than that of the single models according to the experimental outcomes.
Key words : prediction model; time series; prediction of outpatient volume; Extreme Learning Machine(ELM); Empirical Mode Decomposition(EMD)
0 引言
門診工作是現(xiàn)代醫(yī)療工作中非常重要的一環(huán)，同時(shí)日常的門診量也反映著醫(yī)院實(shí)時(shí)的運(yùn)行狀態(tài)，準(zhǔn)確地對(duì)醫(yī)院門診量進(jìn)行有效預(yù)測(cè)，既能為醫(yī)院管理人員進(jìn)行資源合理配置提供重要參考，也能為醫(yī)院的運(yùn)營(yíng)管理起到積極的作用。
門診量預(yù)測(cè)本質(zhì)上是一種時(shí)間序列的預(yù)測(cè)，而大多時(shí)間序列內(nèi)是存在不穩(wěn)定因素的，其中包括就近就醫(yī)、診療質(zhì)量、重點(diǎn)科室知名度、服務(wù)質(zhì)量、就醫(yī)環(huán)境等，這些因素都難以量化。以往研究者對(duì)門診量的預(yù)測(cè)研究只考慮針對(duì)一種或幾種因素，沒(méi)有對(duì)門診量時(shí)間序列數(shù)據(jù)進(jìn)行挖掘，這與深度挖掘技術(shù)在醫(yī)療行業(yè)的研究應(yīng)用較少有關(guān)。
本文詳細(xì)內(nèi)容請(qǐng)下載：http://ihrv.cn/resource/share/2000005381
作者信息：
樊沖
(錦州市大數(shù)據(jù)管理中心,遼寧錦州121000)
此內(nèi)容為AET網(wǎng)站原創(chuàng)，未經(jīng)授權(quán)禁止轉(zhuǎn)載。