Current browse context:
stat.ML
Change to browse by:
References & Citations
Statistics > Machine Learning
Title: Deep Quantile Aggregation
(Submitted on 26 Feb 2021 (this version), latest version 15 Apr 2023 (v5))
Abstract: Conditional quantile estimation is a key statistical learning challenge motivated by the need to quantify uncertainty in predictions or to model a diverse population without being overly reductive. As such, many models have been developed for this problem. Adopting a meta viewpoint, we propose a general framework (inspired by neural network optimization) for aggregating any number of conditional quantile models in order to boost predictive accuracy. We consider weighted ensembling strategies of increasing flexibility where the weights may vary over individual models, quantile levels, and feature values. An appeal of our approach is its portability: we ensure that estimated quantiles at adjacent levels do not cross by applying simple transformations through which gradients can be backpropagated, and this allows us to leverage the modern deep learning toolkit for building quantile ensembles. Our experiments confirm that ensembling can lead to big gains in accuracy, even when the constituent models are themselves powerful and flexible.
Submission history
From: Rasool Fakoor [view email][v1] Fri, 26 Feb 2021 23:21:16 GMT (1952kb,D)
[v2] Tue, 16 Mar 2021 01:23:59 GMT (1952kb,D)
[v3] Fri, 23 Jul 2021 17:11:29 GMT (13661kb,D)
[v4] Thu, 14 Jul 2022 19:03:55 GMT (1930kb,D)
[v5] Sat, 15 Apr 2023 08:40:57 GMT (1985kb,D)
Link back to: arXiv, form interface, contact.