Accurate Uncertainty Estimation and Decomposition in Ensemble Learning

Authors: Jeremiah Liu, John Paisley, Marianthi-Anna Kioumourtzoglou, Brent Coull

NeurIPS 2019 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental This section reports an in-depth validation of the proposed method on a nonlinear function approximation task with complex (heterogeneous and heavy-tailed) observation noise.
Researcher Affiliation Collaboration Jeremiah Zhe Liu Google Research & Harvard University zhl112@mail.harvard.edu John Paisley Columbia University jpaisley@columbia.edu Marianthi-Anna Kioumourtzoglou Columbia University mk3961@cumc.columbia.edu Brent A. Coull Harvard University bcoull@hsph.harvard.edu
Pseudocode No The paper describes methods and processes in narrative text (e.g., Section 2 and Section C) but does not include a formally structured pseudocode or algorithm block.
Open Source Code No The paper does not provide an explicit statement or link indicating that the source code for the described methodology is publicly available.
Open Datasets No We apply BNE to a real-world air pollution prediction ensemble system in Eastern Massachusetts consisted of three state-of-the-art PM2.5 exposure models ([23, 16, 47]). We implement our ensemble framework on the base models out-of-sample predictions at 43 monitors in Eastern Massachusetts in 2011.
Dataset Splits No We vary sample size between 100 and 1000, and repeat the simulation 50 times in each setting.
Hardware Specification No The paper does not provide any specific hardware details (such as GPU/CPU models, memory, or specific computing environments) used for running its experiments.
Software Dependencies No Posterior sampling is performed using Hamiltonian Monte Carlo (HMC) [2]
Experiment Setup No The detailed experiment settings are documented in Supplementary G.