摘要
将分位数回归理论引入到回归模型中,应用线性条件分位数回归模型分析海平面高度与时间及南方涛动指数之间的线性变化趋势,得到参数估计值。结果表明:年最高海平面高度在不同的分位数下与时间及南方涛动指数变化趋势是不同的,而由最小二乘法回归只能得到平均变化趋势。分位数回归比经典的最小二乘回归能够反映更多的局部信息。
The theory of quantile regression is introduced to regression model. Linear conditional quantile regression model is applied to analyze the linear trend of the sea level related with time and Southern Oscillation Index. The value of parametric estimation is calculated. The result shows that the linear trend of the annual maximum sea level related with time and Southern Oscillation Index is different under different quantiles. Quantile regression can provide much more information than the classical least square regression.
出处
《海洋技术》
北大核心
2009年第3期83-86,共4页
Ocean Technology
基金
国家自然科学基金资助项目(70573077)
关键词
分位数回归
海平面高度
南方涛动指数
参数估计
quantile regression
sea level
Southern Oscillation Index
parametric estimation
