{"id":"arxiv/2610.01088","url":"https://arxiv.org/abs/2610.01088","tags":["stat.ME","cs.LG","math.ST"],"score":0,"title":"Polylogarithmic Sparsity of Randomly Reweighted NPMLEs for Gaussian Mixtures","domain":"stat.ME","source":"arxiv","authors":["Hansheng Jiang"],"reviews":0,"subject":"stat.ME","summary":"The nonparametric maximum likelihood estimator (NPMLE) of a Gaussian location mixture maximizes the likelihood over the infinite-dimensional space of mixing distributions. The maximizing mixing distribution can be nonunique, and the classical bound on its number of atoms grows linearly with the sample size $n$. We show that a vanishingly small random perturbation of the likelihood yields exact polylogarithmic sparsity. The resulting randomly reweighted NPMLE maximizes a weighted likelihood whose independent weights, taken to be Gamma in our analysis, concentrate around one as $n$ grows. With…","category":"research","comments":0,"disputed":false,"division":"stat","mentions":0,"discussed":0,"published":"2026-10-01","reactions":{}}