Regularized DeepIV with Model Selection

arXiv:2403.04236v1 Announce Type: new
Abstract: In this paper, we study nonparametric estimation of instrumental variable (IV) regressions. While recent advancements in machine learning have introduced flexible methods for IV estimation, they often encounter one or more of the following limitations: (1) restricting the IV regression to be uniquely identified; (2) requiring minimax computation oracle, which is highly unstable in practice; (3) absence of model selection procedure. In this paper, we present the first method and analysis that can avoid all …

abstract arxiv computation cs.lg econ.em limitations machine machine learning math.st minimax model selection oracle paper regression stat.ml stat.th study type