本页提供题录、可核实摘要与阅读线索。AI 解读未完成全文核验，可能有误；具体假设、证明和结论请核对原文。仅依据书目生成的内容属于阅读导航，不是对论文结论的确认。
摘要保留来源文本，未经模型改写。材料不足时保留待解读状态；引用采用 OpenAlex 单条关联记录，不合计版本，缺失不视为零，也不代表论文质量。
正在核对对应版本的资源…
用完备性刻画非参数条件矩限制的识别，连接识别与逆问题。
In econometrics there are many occasions where knowledge of the structural relationship among dependent variables is required to answer questions of interest. This paper gives identification and estimation results for nonparametric conditional moment restrictions. We characterize identification of structural functions as completeness of certain conditional distributions, and give sufficient identification conditions for exponential families and discrete variables. We also give a consistent, nonparametric estimator of the structural function. The estimator is nonparametric two-stage least squares based on series approximation, which overcomes an ill-posed inverse problem by placing bounds on integrals of higher-order derivatives.