Model orthogonalization and Bayesian forecast mixing via Principal Component Analysis

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Giuliani, Pablo, Godbey, Kyle, Kejzlar, Vojtech, Nazarewicz, Witold
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866910570843734016
author Giuliani, Pablo
Godbey, Kyle
Kejzlar, Vojtech
Nazarewicz, Witold
author_facet Giuliani, Pablo
Godbey, Kyle
Kejzlar, Vojtech
Nazarewicz, Witold
contents One can improve predictability in the unknown domain by combining forecasts of imperfect complex computational models using a Bayesian statistical machine learning framework. In many cases, however, the models used in the mixing process are similar. In addition to contaminating the model space, the existence of such similar, or even redundant, models during the multimodeling process can result in misinterpretation of results and deterioration of predictive performance. In this work we describe a method based on the Principal Component Analysis that eliminates model redundancy. We show that by adding model orthogonalization to the proposed Bayesian Model Combination framework, one can arrive at better prediction accuracy and reach excellent uncertainty quantification performance.
format Preprint
id arxiv_https___arxiv_org_abs_2405_10839
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Model orthogonalization and Bayesian forecast mixing via Principal Component Analysis
Giuliani, Pablo
Godbey, Kyle
Kejzlar, Vojtech
Nazarewicz, Witold
Nuclear Theory
Data Analysis, Statistics and Probability
Machine Learning
One can improve predictability in the unknown domain by combining forecasts of imperfect complex computational models using a Bayesian statistical machine learning framework. In many cases, however, the models used in the mixing process are similar. In addition to contaminating the model space, the existence of such similar, or even redundant, models during the multimodeling process can result in misinterpretation of results and deterioration of predictive performance. In this work we describe a method based on the Principal Component Analysis that eliminates model redundancy. We show that by adding model orthogonalization to the proposed Bayesian Model Combination framework, one can arrive at better prediction accuracy and reach excellent uncertainty quantification performance.
title Model orthogonalization and Bayesian forecast mixing via Principal Component Analysis
topic Nuclear Theory
Data Analysis, Statistics and Probability
Machine Learning
url https://arxiv.org/abs/2405.10839