组合预测法在我国汽车市场需求预测中的应用

The Application of Combinatorial Forecast on Chinese Automobile Market Demand

组合预测理论是对同一个预测对象采用不同的单项预测模型,并对各个单项模型施以适当权重。因而,利用多个单项预测模型有效组合更多有用的信息资源,使组合预测模型具有较高的预测精度和预测稳定性,能比较合理地描述系统的客观现实。依据中国民用汽车保有量1985-2003年的历史数据及汽车市场需求系统的复杂特性,分别采用灰色系统、多元回归、三次指数平滑方法建立单项预测模型。根据组合预测理论建立中国汽车市场需求组合预测模型;运用组合预测模型对我国汽车市场的未来需求进行了预测。

Combinatorial forecasting theory involves different single forecasting models for one forecasting object, and accords proper authority to every single forecasting model. With these single forecasting models to effeetively assemble useful information resource, combinatorial forecasting model can be more precise and stable.This helps to objectively represent the truth of system. According to the domestic automobile retaining quantity sampled from 1985 to 2003 and the complexity of automobile market demand systems, the paper presents grey system, multiple regression, thrice index smoothness to establish single forecasting model. Based on combinatorial forecasting theory, the future demand of automobile market in China can be more precisely forecasted.