Current browse context:
cs.LG
Change to browse by:
References & Citations
Computer Science > Machine Learning
Title: Learning Similarity Metrics for Numerical Simulations
(Submitted on 18 Feb 2020 (v1), last revised 23 Jun 2020 (this version, v2))
Abstract: We propose a neural network-based approach that computes a stable and generalizing metric (LSiM) to compare data from a variety of numerical simulation sources. We focus on scalar time-dependent 2D data that commonly arises from motion and transport-based partial differential equations (PDEs). Our method employs a Siamese network architecture that is motivated by the mathematical properties of a metric. We leverage a controllable data generation setup with PDE solvers to create increasingly different outputs from a reference simulation in a controlled environment. A central component of our learned metric is a specialized loss function that introduces knowledge about the correlation between single data samples into the training process. To demonstrate that the proposed approach outperforms existing metrics for vector spaces and other learned, image-based metrics, we evaluate the different methods on a large range of test data. Additionally, we analyze generalization benefits of an adjustable training data difficulty and demonstrate the robustness of LSiM via an evaluation on three real-world data sets.
Submission history
From: Georg Kohl [view email][v1] Tue, 18 Feb 2020 20:11:15 GMT (5342kb,D)
[v2] Tue, 23 Jun 2020 18:14:55 GMT (5078kb,D)
Link back to: arXiv, form interface, contact.