{"id":"arxiv/2603.14037","url":"https://arxiv.org/abs/2603.14037","tags":["math.ST"],"score":0,"title":"On a Universal Strictly Decreasing Nonparametric Estimator Applied to the Drift Function of a Recurrent Diffusion Process Estimation","domain":"math.ST","source":"arxiv","authors":["Nicolas Marie"],"reviews":0,"subject":"math.ST","summary":"This paper deals with a copies-based continuously differentiable and strictly decreasing estimator of the drift function for stochastic differential equations defining recurrent diffusion processes. The first part of our paper deals with non-asymptotic $\\mathbb L^1$-risk bounds and a bandwidths selection procedure for a universal monotone estimator. These results are tailor-made to our framework, and then applied to the estimation of the drift function of recurrent diffusion processes in the second part of the paper.","category":"research","comments":0,"disputed":false,"division":"math","mentions":0,"discussed":0,"published":"2026-03-14","reactions":{}}