stat.MLNov 13, 2026

Taking Band-Limited Maximum Likelihood to High Dimensions: Linear Memory, the Parametric Rate, and a Self-Tuning Cut-off

Authors: Rahul Agarwal, Ankit Agarwal

Organizations: Independent researcher (Ph.D., Biomedical Engineering, Johns Hopkins University) Sammamish, WA, USA · Independent researcher (M.B.A., University of Southern California) San Jose, CA, USA

Abstract

Band-limited maximum likelihood (BLML) estimates a density as the square of a band-limited function. The estimate is non-negative, and when the square root of the true density is band-limited its error falls at the parametric rate n−1n^{-1}. Its algorithm bins the data, which is fast only in one or two dimensions: from three on every point has its own bin, the kernel system is dense, and memory grows as n2n^2, 40 GB at 10510^5 points. Whether the rate survives in higher dimension, where kernel estimators slow down, had not been tested. We study dense data, densities that fill their space, not data near a low-dimensional manifold. We show that the binned equations are the stationarity condition of a strictly convex objective and minimise it by block coordinate descent without forming the kernel matrix: memory is linear in nn, a sweep costs O(n2d)O(n^2 d) and a fit takes about ten, and one A100 GPU fits 10610^6 points in eight dimensions in about ten minutes. On band-limited truths the error then decays as n−1n^{-1} in every dimension we tested, up to eight, while the slopes of kernel estimators flatten from -0.67 to -0.28, and it is 2 to 18 times lower than a neural spline flow's. A streaming fit learns the cut-off, BLML's only tuning parameter, from analytic gradients of the held-out likelihood, for about the cost of one more fit. On real data BLML is the strongest nonparametric estimator we tested: it leads kernel and nearest-neighbour estimators on electricity prices by 0.27 to 0.47 nats per hour, and a tuned kernel estimator on the full POWER and GAS benchmarks by 0.39 and 1.58 nats per point, placing it between early and recent normalizing flows. It loses on densities with a kink.

Figures & tables

Appendix figures & tables16 assets

Supplementary material from the paper’s appendix.

Appendix

Explore similar work

CardsList
  1. Breaking the Curse with BAND: Nonparametric Distribution Estimation in High Dimensions

    Jul 29, 2026Shuo-Chieh Huang, Chien-Ming Chi, Jau-er ChenBayesian NetworksDensity Estimation

  2. Adaptive Kernel Density Estimation with Pre-training

    May 13, 2026Ruitong Zhang, Ke DengKernel Density EstimationDensity Estimation

  3. Exact Schur-Sylvester Dimensionality Reductions for Non-Smooth Stochastic Complexity and Manifold Sampling

    Jun 22, 2026Trenton Lau, Gary P. T. ChoiMarkov Chain Monte Carlo