Discriminative learning of prediction intervals

Nir Rosenfeld, Yishay Mansour, Elad Yom-Tov

Research output: Contribution to conferencePaperpeer-review

Abstract

In this work we consider the task of constructing prediction intervals in an inductive batch setting. We present a discriminative learning framework which optimizes the expected error rate under a budget constraint on the interval sizes. Most current methods for constructing prediction intervals offer guarantees for a single new test point. Applying these methods to multiple test points can result in a high computational overhead and degraded statistical guarantees. By focusing on expected errors, our method allows for variability in the per-example conditional error rates. As we demonstrate both analytically and empirically, this flexibility can increase the overall accuracy, or alternatively, reduce the average interval size. While the problem we consider is of a regressive flavor, the loss we use is combinatorial. This allows us to provide PAC-style, finite-sample guarantees. Computationally, we show that our original objective is NP-hard, and suggest a tractable convex surrogate. We conclude with a series of experimental evaluations.

Original languageEnglish
Pages347-355
Number of pages9
StatePublished - 2018
Event21st International Conference on Artificial Intelligence and Statistics, AISTATS 2018 - Playa Blanca, Lanzarote, Canary Islands, Spain
Duration: 9 Apr 201811 Apr 2018

Conference

Conference21st International Conference on Artificial Intelligence and Statistics, AISTATS 2018
Country/TerritorySpain
CityPlaya Blanca, Lanzarote, Canary Islands
Period9/04/1811/04/18

Fingerprint

Dive into the research topics of 'Discriminative learning of prediction intervals'. Together they form a unique fingerprint.

Cite this