Filling the Void: Data-Driven Machine Learning-based Reconstruction of Sampled Spatiotemporal Scientific Simulation Data

As high-performance computing systems continue to advance, the gap between computing performance and I/O capabilities is widening. This bottleneck limits the storage capabilities of increasingly large-scale simulations, which generate data at never-before-seen granularities while only being able to...

Full description

Saved in:
Bibliographic Details
Published in:SC24-W: Workshops of the International Conference for High Performance Computing, Networking, Storage and Analysis pp. 290 - 299
Main Authors: Biswas, Ayan, Mishra, Aditi, Majumder, Meghanto, Hazarika, Subhashis, Most, Alexander, Castorena, Juan, Bryan, Christopher, McCormick, Patrick, Ahrens, James, Lawrence, Earl, Hagberg, Aric
Format: Conference Proceeding
Language:English
Published: IEEE 17.11.2024
Subjects:
Computational modeling
computing methodologies
Data models
data reduction
Deep learning
Exascale computing
Filling
Interpolation
Knowledge transfer
machine learning
machine learning approaches
modeling and simulation
Neural networks
reconstruction
scientific visualization
simulation types and techniques
Spatiotemporal phenomena
Training
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:As high-performance computing systems continue to advance, the gap between computing performance and I/O capabilities is widening. This bottleneck limits the storage capabilities of increasingly large-scale simulations, which generate data at never-before-seen granularities while only being able to store a small subset of the raw data. Recently, strategies for data-driven sampling have been proposed to intelligently sample the data in a way that achieves high data reduction rates while preserving important regions or features with high fidelity. However, a thorough analysis of how such intelligent samples can be used for data reconstruction is lacking. We propose a data-driven machine learning approach based on training neural networks to reconstruct full-scale datasets based on a simulation's sampled output. Compared to current state-of-the-art reconstruction approaches such as Delaunay triangulation-based linear interpolation, we demonstrate that our machine learning-based reconstruction has several advantages, including reconstruction quality, time-to-reconstruct, and knowledge transfer to unseen timesteps and grid resolutions. We propose and evaluate strategies that balance the sampling rates with model training (pretraining and fine-tuning) and data reconstruction time to demonstrate how such machine learning approaches can be tailored for both speed and quality for the reconstruction of grid-based datasets.
DOI:10.1109/SCW63240.2024.00045