小波分析和支持向量机组合法预测应急血液需求研究
|
摘要
针对地震紧急救援阶段血液需求特点,提出用于预测其需求量的,基于小波分析(WA)和支持向量机(SVM)的组合方法(WA-SVM)。首先对原始血液需求进行小波分析,然后确定SVM输入向量和输出向量集合,构建各层序列的SVM预测模型,对血液需求进行预测。汶川地震案例表明,该方法的预测精度优于经验模态分解和SVM的组合预测模型以及SVM单项预测模型。
A combined method of WA and SVM was worked out for forecasting the blood demand in emergency rescue after earthquake.The sample demand series was first mapped into several different frequency scales using wavelet transformation,and then a SVM is established for each scale.The WA-SVM forecasting method was then used to forecast blood demand in emergency rescue taking a case in Wenchuan earthquake as an example.The forecasting results indicate that the WA-SVM forecasting method has better performance both in accuracy and applicability in comparison with the empirical mode decomposition(EMD)-SVM combined method,and SVM method.
A combined method of WA and SVM was worked out for forecasting the blood demand in emergency rescue after earthquake.The sample demand series was first mapped into several different frequency scales using wavelet transformation,and then a SVM is established for each scale.The WA-SVM forecasting method was then used to forecast blood demand in emergency rescue taking a case in Wenchuan earthquake as an example.The forecasting results indicate that the WA-SVM forecasting method has better performance both in accuracy and applicability in comparison with the empirical mode decomposition(EMD)-SVM combined method,and SVM method.
引文
[1]Hrezo R J,Clark J.The walking blood bank:An alternative blood supply in military mass casualties[J].DisasterManagement&Response,2003,1(1):19-22.
[2]Tabatabaie M,Ardalan A,Abolghasemi H,et al.Estimating blood transfusion requirements in preparation for a majorearthquake:the Tehran,Iran study[J].Prehosp Disaster Med,2010,25(3):246-252.
[3]Contreras J,Esipinola R,Nogales F J,et al.ARIMA models to predict next-day electricity prices[J].IEEE Transactionson Power Systems,2003,18(3):1014-1020.
[4]Gutierrez R S,Solis A O,Mukhopadhyay S.Lumpy demand forecasting using neural networks[J].International Journalof Production Economics,2008,111(2):409-420.
[5]XU X H,ZHANG H.Forecasting demand of short life cycle products by SVM[C].2008 International Conference onManagement Science&Engineering,2008:352-356.
[6]LIU K,Subbarayan S,Shoults R R,et al.Comparison of very short-term load forecasting techniques[J].IEEE Transac-tions on Power Systems,1996,11(2):877-882.
[7]XU X Y,QI Y Q,Hua Z S.Forecasting demand of commodities after natural disasters[J].Expert System with Applica-tions,2010,37(2010):4 313-4 317.
[8]Vapnik V,Golowich SE,Smola A.Support vector method for function approximation,regression estimation,and signalprocessing[C].Advances in Neural Information Processing Systems 9.Proceedings of the 1996 Neural InformationProcessing Systems Conference,1996:281-287.
[9]Mukherjee S,Osuna E,Girosi F.Nonlinear prediction of chaotic time series using support vector machines[C].NeuralNetworks for Signal Processing VII.Proceeding of the 1997 IEEE Workshop,1997:536-542.
[10]D H,ZHAO X Y,LIU Y,et al.Stochastic traffic series time combination forecasting based on wavelet analysis[J].Journal of System Simulation,2012,24(2):377-381.
[11]程勤侦,田伟,赵晓嘉.特大地震灾害期间医院血液需求的变化及其影响因素分析[J].中国输血杂志,2008,21(8):588-591.CHENG Qin-zhen,TIAN Wei,ZHAO Xiao-jia.Study on the changes of blood demand and the related factors during theWenchuan earthquake relief work[J].Chinese Journal of Blood Transfusion,2008,21(8):588-591.
[2]Tabatabaie M,Ardalan A,Abolghasemi H,et al.Estimating blood transfusion requirements in preparation for a majorearthquake:the Tehran,Iran study[J].Prehosp Disaster Med,2010,25(3):246-252.
[3]Contreras J,Esipinola R,Nogales F J,et al.ARIMA models to predict next-day electricity prices[J].IEEE Transactionson Power Systems,2003,18(3):1014-1020.
[4]Gutierrez R S,Solis A O,Mukhopadhyay S.Lumpy demand forecasting using neural networks[J].International Journalof Production Economics,2008,111(2):409-420.
[5]XU X H,ZHANG H.Forecasting demand of short life cycle products by SVM[C].2008 International Conference onManagement Science&Engineering,2008:352-356.
[6]LIU K,Subbarayan S,Shoults R R,et al.Comparison of very short-term load forecasting techniques[J].IEEE Transac-tions on Power Systems,1996,11(2):877-882.
[7]XU X Y,QI Y Q,Hua Z S.Forecasting demand of commodities after natural disasters[J].Expert System with Applica-tions,2010,37(2010):4 313-4 317.
[8]Vapnik V,Golowich SE,Smola A.Support vector method for function approximation,regression estimation,and signalprocessing[C].Advances in Neural Information Processing Systems 9.Proceedings of the 1996 Neural InformationProcessing Systems Conference,1996:281-287.
[9]Mukherjee S,Osuna E,Girosi F.Nonlinear prediction of chaotic time series using support vector machines[C].NeuralNetworks for Signal Processing VII.Proceeding of the 1997 IEEE Workshop,1997:536-542.
[10]D H,ZHAO X Y,LIU Y,et al.Stochastic traffic series time combination forecasting based on wavelet analysis[J].Journal of System Simulation,2012,24(2):377-381.
[11]程勤侦,田伟,赵晓嘉.特大地震灾害期间医院血液需求的变化及其影响因素分析[J].中国输血杂志,2008,21(8):588-591.CHENG Qin-zhen,TIAN Wei,ZHAO Xiao-jia.Study on the changes of blood demand and the related factors during theWenchuan earthquake relief work[J].Chinese Journal of Blood Transfusion,2008,21(8):588-591.
版权所有：© 2023 中国地质图书馆 中国地质调查局地学文献中心