Huggins Richard M, Hall Peter, Yip Paul S F, Bui Quang M
Centre for Mathematics and its Applications, Mathematical Sciences Institute, Australian National University, Canberra 0200 ACT, Australia.
Biometrics. 2007 Sep;63(3):708-13. doi: 10.1111/j.1541-0420.2006.00727.x.
A semivarying coefficient model for the monthly numbers of suicides in Hong Kong is developed and a new estimation procedure for estimating the parametric component is proposed. The estimators are examined in a small simulation study and fitted to monthly suicide data to estimate a nonparametric long-term trend and parametric seasonal and socioeconomic effects. Fitting the model detected interpretable structure in the data that is consistent with that driving public health policy. While exploratory, the analysis motivates the collection of more detailed data and the development of more sophisticated models to help determine target groups and strategies to reduce the suicide rate in Hong Kong.
本文建立了一个用于分析香港每月自杀人数的半可变系数模型，并提出了一种估计参数成分的新估计程序。通过一个小型模拟研究对估计量进行了检验，并将其应用于每月自杀数据，以估计非参数长期趋势以及参数化的季节和社会经济效应。对该模型进行拟合后，发现数据中存在可解释的结构，这与推动公共卫生政策的因素相一致。虽然该分析具有探索性，但它促使人们收集更详细的数据，并开发更复杂的模型，以帮助确定目标群体和降低香港自杀率的策略。
