{"id":"arxiv/2607.24472","url":"https://arxiv.org/abs/2607.24472","tags":["econ.EM","math.ST","stat.ME","stat.ML"],"score":0,"title":"Debiased Machine Learning: Identification, Estimation, and Shape Constraints","domain":"econ.EM","source":"arxiv","authors":["Qihui Chen","Ka Yan Cheng","Zheng Fang"],"reviews":0,"subject":"econ.EM","summary":"We develop a general framework of identification and estimation for automatic debiased machine learning (DML) where the parameter of interest $θ_0$ is identified by a moment condition involving a nuisance $γ_0$ that may be high dimensional. We establish conditions under which the Riesz representer $α_0$, which is at the core of DML, is identified, and show that the identification occurs precisely when $α_0$ uniquely optimizes a quadratic functional. This characterization enables us to develop a general estimation procedure for $α_0$ that allows for generic $γ_0$ including those defined by…","category":"research","comments":0,"disputed":false,"division":"econ","mentions":0,"discussed":0,"published":"2026-07-27","reactions":{}}