鞅差误差下部分函数线性模型的经验似然推断

Empirical Likelihood Inference for Partial Functional Linear Models under Martingale Difference Sequence

研究了误差是鞅差序列下的部分函数线性回归模型的经验似然估计问题.首先构造了回归系数的经验似对数似然比统计量,进一步证明了所提出的估计量渐近服从卡方分布.利用这一结果可以用来构造相应参数的置信域.

This paper, studies the empirical likelihood estimation for partial functional linear models under martingale difference sequence.First of all,the regression coefficient of experience logarithmic likelihood ratio statistics was constructed and it was proven that the proposed estimator asymptotically obeys the chi-square distribution.The results can be used to construct the corresponding parameter confidence regions.