We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:

Download:

Current browse context:

stat.ML

Change to browse by:

References & Citations

Bookmark

(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Statistics > Machine Learning

Title: Universal Prediction Band via Semi-Definite Programming

Abstract: We propose a computationally efficient method to construct nonparametric, heteroskedastic prediction bands for uncertainty quantification, with or without any user-specified predictive model. The data-adaptive prediction band is universally applicable with minimal distributional assumptions, with strong non-asymptotic coverage properties, and easy to implement using standard convex programs. Our approach can be viewed as a novel variance interpolation with confidence and further leverages techniques from semi-definite programming and sum-of-squares optimization. Theoretical and numerical performances for the proposed approach for uncertainty quantification are analyzed.
Comments: 15 pages, 3 figures
Subjects: Machine Learning (stat.ML); Machine Learning (cs.LG); Econometrics (econ.EM); Optimization and Control (math.OC); Statistics Theory (math.ST)
Cite as: arXiv:2103.17203 [stat.ML]
  (or arXiv:2103.17203v1 [stat.ML] for this version)

Submission history

From: Tengyuan Liang [view email]
[v1] Wed, 31 Mar 2021 16:30:58 GMT (1194kb,D)

Link back to: arXiv, form interface, contact.