从回归方程可知: 焦碳价格对生铁价格的线性回归结果要好于对焦碳产量线性回归结果。所以, 焦炭价格短期预测的半参数模型的参数部分选生铁价格变量, 非参数部分选焦碳产量变量
上海翻译公司。
It can be known from the regression equation that the linear regression results of the coke price against the cast iron price is better than against the linear regression results of coke output. So, cast iron is a variable among the parameters of the semi-parametric model of the short-term coke price
半参数—误差修正模型估计拟合图
Semi-parametric—error correction model estimate fit chart
把ΔY
t的计算结果带入到公式(1)中, 核函数选取抛物线核, 通过
交错鉴定法求解到最佳
窗宽h=0.298, 半参数模型的估计方法选择最小二乘核估计法
【16-20】。图5是TAN40短期价格的半参数——误差修正法的估计拟合图。图6是短期焦炭价格的半参数非线性(二次多项式)估计的拟合图
北京翻译公司。
Substitute the arithmetic operation result of ΔYt in the formula (1). In the function, parabola kernel is chosen to work out, on the basis of cross identification method, the maximum window width of h=0.298. In the semi-parametric model, the least square estimate method【16-20】 is applied. The Chart 5 is a fit chart for the non-linear semi-parametric estimate (
quadratic polynomial) of the short-term coke price.