Multi-task Modeling for Engineering Applications with Sparse Data

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Comlek, Yigitcan, Krishnan, R. Murali, Ravi, Sandipp Krishnan, Moghaddas, Amin, Giorjao, Rafael, Eff, Michael, Samaddar, Anirban, Ramachandra, Nesar S., Madireddy, Sandeep, Wang, Liping
Format: Preprint
Published: 2026
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866914243443425280
author Comlek, Yigitcan
Krishnan, R. Murali
Ravi, Sandipp Krishnan
Moghaddas, Amin
Giorjao, Rafael
Eff, Michael
Samaddar, Anirban
Ramachandra, Nesar S.
Madireddy, Sandeep
Wang, Liping
author_facet Comlek, Yigitcan
Krishnan, R. Murali
Ravi, Sandipp Krishnan
Moghaddas, Amin
Giorjao, Rafael
Eff, Michael
Samaddar, Anirban
Ramachandra, Nesar S.
Madireddy, Sandeep
Wang, Liping
contents Modern engineering and scientific workflows often require simultaneous predictions across related tasks and fidelity levels, where high-fidelity data is scarce and expensive, while low-fidelity data is more abundant. This paper introduces an Multi-Task Gaussian Processes (MTGP) framework tailored for engineering systems characterized by multi-source, multi-fidelity data, addressing challenges of data sparsity and varying task correlations. The proposed framework leverages inter-task relationships across outputs and fidelity levels to improve predictive performance and reduce computational costs. The framework is validated across three representative scenarios: Forrester function benchmark, 3D ellipsoidal void modeling, and friction-stir welding. By quantifying and leveraging inter-task relationships, the proposed MTGP framework offers a robust and scalable solution for predictive modeling in domains with significant computational and experimental costs, supporting informed decision-making and efficient resource utilization.
format Preprint
id arxiv_https___arxiv_org_abs_2601_05910
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Multi-task Modeling for Engineering Applications with Sparse Data
Comlek, Yigitcan
Krishnan, R. Murali
Ravi, Sandipp Krishnan
Moghaddas, Amin
Giorjao, Rafael
Eff, Michael
Samaddar, Anirban
Ramachandra, Nesar S.
Madireddy, Sandeep
Wang, Liping
Machine Learning
Applications
Modern engineering and scientific workflows often require simultaneous predictions across related tasks and fidelity levels, where high-fidelity data is scarce and expensive, while low-fidelity data is more abundant. This paper introduces an Multi-Task Gaussian Processes (MTGP) framework tailored for engineering systems characterized by multi-source, multi-fidelity data, addressing challenges of data sparsity and varying task correlations. The proposed framework leverages inter-task relationships across outputs and fidelity levels to improve predictive performance and reduce computational costs. The framework is validated across three representative scenarios: Forrester function benchmark, 3D ellipsoidal void modeling, and friction-stir welding. By quantifying and leveraging inter-task relationships, the proposed MTGP framework offers a robust and scalable solution for predictive modeling in domains with significant computational and experimental costs, supporting informed decision-making and efficient resource utilization.
title Multi-task Modeling for Engineering Applications with Sparse Data
topic Machine Learning
Applications
url https://arxiv.org/abs/2601.05910