MelissaDL x Breed: Towards Data-Efficient On-line Supervised Training of Multi-parametric Surrogates with Active Learning
Not on the map. Why?
Abstract (arXiv)
Artificial intelligence is transforming scientific computing with deep neural network surrogates that approximate solutions to partial differential equations (PDEs). Traditional off-line training methods face issues with storage and I/O efficiency, as the training dataset has to be computed with numerical solvers up-front. Our previous work, the Melissa framework, addresses these problems by enabling data to be created "on-the-fly" and streamed directly into the training process. In this paper we introduce a new active learning method to enhance data-efficiency for on-line surrogate training. The surrogate is direct and multi-parametric, i.e., it is trained to predict a given timestep directly with different initial and boundary conditions parameters. Our approach uses Adaptive Multiple Importance Sampling guided by training loss statistics, in order to focus NN training on the difficult areas of the parameter space. Preliminary results for 2D heat PDE demonstrate the potential of this method, called Breed, to improve the generalization capabilities of surrogates while reducing computational overhead.
Key facts. Cited by 1 paper in the map (data through 14 Oct 2026). Builds on 31 of its 46 references that are in the map. First posted on arXiv 8 Oct 2024.
Builds on
31 of 46 references.
References · 31
- 2020Fourier Neural Operator for Parametric Partial Differential Equationscited by 2,908
- 2020Learning Mesh-Based Simulation with Graph Networkscited by 868
- 2020A Survey of Deep Active Learningcited by 554
- 2019The frontier of simulation-based inferencecited by 502
- 2023ClimaX: A foundation model for weather and climatecited by 337
- 2018Not All Samples Are Created Equal: Deep Learning with Importance Samplingcited by 328
- 2023Neural General Circulation Models for Weather and Climatecited by 239
- 2022A comprehensive study of non-adaptive and residual-based adaptive sampling for physics-informed neural networkscited by 236
- 2023Neural Operators for Accelerating Scientific Simulations and Designcited by 221
- 2024Poseidon: Efficient Foundation Models for PDEscited by 200
- 2023An Expert's Guide to Training Physics-informed Neural Networkscited by 126
- 2021Efficient training of physics-informed neural networks via importance samplingcited by 110
19 more
- 2022Mitigating Propagation Failures in Physics-informed Neural Networks using Retain-Resample-Release (R3) Samplingcited by 73
- 2023Benchmarking Autoregressive Conditional Diffusion Models for Turbulent Flow Simulationcited by 71
- 2021Simulation Intelligence: Towards a New Generation of Scientific Methodscited by 68
- 2021Revisiting Methods for Finding Influential Examplescited by 39
- 2021Deep Active Learning by Leveraging Training Dynamicscited by 32
- 2024PINNACLE: PINN Adaptive ColLocation and Experimental points selectioncited by 32
- 2023Active learning for data streams: a surveycited by 29
- 2022AUTOMATA: Gradient Based Data Subset Selection for Compute-Efficient Hyper-parameter Tuningcited by 22
- 2024Active Learning for Neural PDE Solverscited by 22
- 2019Etalumis: Bringing Probabilistic Programming to Scientific Simulators at Scalecited by 18
- 2022BERT on a Data Diet: Finding Important Examples by Gradient-Based Pruningcited by 17
- 2022TiDAL: Learning Training Dynamics for Active Learningcited by 11
- 2021Colmena: Scalable Machine-Learning-Based Steering of Ensemble Simulations for High Performance Computingcited by 10
- 2022Deep Active Learning for Computer Vision: Past and Futurecited by 7
- 2023Training Deep Surrogate Models with Large Scale Online Learningcited by 5
- 2024Reference Neural Operators: Learning the Smooth Dependence of Solutions of PDEs on Geometric Deformationscited by 5
- 2024Provably Neural Active Learning Succeeds via Prioritizing Perplexing Samplescited by 5
- 2023High Throughput Training of Deep Surrogates from Large Ensemble Runscited by 3
- 2023Population Monte Carlo with Normalizing Flowcited by 1
Built on by
Citing papers per year: 2026: 1.
1 paper.