灰色马尔科夫模型在湖北新冠肺炎出院人数预测中的应用

The Application of Grey-Markov Model in Forecasting the Number of Discharged Patients with Novel Coronavirus Pneumonia in Hubei Province

目的建立灰色马尔科夫预测模型,对湖北省新冠肺炎每日出院人数进行预测。方法基于2020年3月1日~3月21日的湖北省卫生健康委员会官网新冠肺炎疫情直报数据,构建灰色预测模型,马尔科夫模型修正预测结果,拟对3月22日~3月24日的湖北省新冠肺炎每日出院人数进行预测。结果2020年3月22日~3月24日的灰色预测值分别为635、594、555,而实际值分别为490、447、444,其平均相对误差为29.16%。修正后,修正值分别为556、520、486,其平均相对误差为13.02%,误差降低55.17%。灰色预测模型的C=0.30599,P=3.5,灰色马尔科夫模型的C=0.10699,P=3.5。结论灰色马尔科夫模型的精确度高于单一的灰色预测模型,能够较好地预测每日出院人数。

Objective To predict novel coronavirus pneumonia discharged in Hubei province by setting up grey-markov model.Methods The grey-marko model was setup by using the data of novel coronavirus pneumonia reported on the official website of hubei provincial health commission from March 1 to 21,then used to predict the number of newly discharged patients with novel coronavirus pneumonia in Hubei province of March 22 to 24.Results The grey prediction values are 635,594 and 555 respectively from March 22 to 24 while the actual values are 490,447 and 444 respectively,with an average relative error of 29.16%.After the correction,the correction values are 556,520 and 486 respectively,the average relative error is 13.02%,and the error is reduced by 55.17%.C=0.30599,P=3.5 for the grey model,and C=0.10699,P=3.5 for the grey markov model.Conclusion The prediction accuracy of grey markov model is higher than that of grey model.The grey and marko model can be used to predict the number of newly discharged patients with high prediction precision.