论文部分内容阅读
本文在卫生统计领域中引入时序方法对人口死亡数(包括各种死因死亡数)建立了不同的时序模型(AR 模型、多维 AR 模型和 SETAR 模型),并用模型对人口死亡数作群体预报和气象因素分析,结果令人满意。此项工作表明,在卫生统计领域中引入时序方法将是进行死亡预报和因素分析的一条新途径。人口死亡的统计、预报和分析是卫生统计部门的一项重要工作。我们根据我市某区防疫站提供的群体人口死亡资料及肿瘤、心脏病等死因死亡资料,应用时间序列分析方法建立了 AR 模型、多维 AR 模型和自激励门限自回归模型(Self-Excited Threshold Auto-Regreesive Model,简记为 SETAR模型),并用这些模型对人口死亡情况进行群体预报和气象因素分析。由于 AR 模型、多维 AR 模型已逐步为人们所熟知,本文将首先着重论述 SETAR 模型的意义及建模方法,然后分别示出这三种模型的预报及分析结果。
In the field of health statistics, a time series method is introduced to establish different time series models (AR model, multidimensional AR model and SETAR model) for the number of deaths of the population (including deaths of all kinds of deaths), and the model is used to make population forecasts of population deaths and meteorological Factor analysis, the results are satisfactory. This work shows that the introduction of timing methods in the field of health statistics will be a new way to predict mortality and factor analysis. Statistics, forecasts and analyzes of the death of a population are an important part of health statistics. Based on the death data of population and deaths of patients such as cancer and heart disease provided by the epidemic prevention station in a certain area of our city, we established the AR model, the multi-dimensional AR model and the self-excitation threshold auto-regression model (Self-Excited Threshold Auto -Regreesive Model, abbreviated as SETAR model), and use these models to make population forecast and weather factor analysis on the death of population. Due to the AR model and multidimensional AR model have gradually been well known, this paper will focus on the significance of the SETAR model and modeling methods, and then show the prediction and analysis results of these three models respectively.