The DOI matches its Crossref record; 6 of the 6 references it cites were found in a registry or an index.
Adaptive activation functions accelerate convergence in deep and physics-informed neural networks
Open accessAbstract
Abstract We employ adaptive activation functions for regression in deep and physics-informed neural networks (PINNs) to approximate smooth and discontinuous functions as well as solutions of linear and nonlinear partial differential equations. In particular, we solve the nonlinear Klein-Gordon equation, which has smooth solutions, the nonlinear Burgers equation, which can admit high gradient solutions, and the Helmholtz equation. We introduce a scalable hyper-parameter in the activation function, which can be optimized to achieve best performance of the network as it changes dynamically the topology of the loss function involved in the optimization process. The adaptive activation function has better learning capabilities than the traditional one (fixed activation) as it improves greatly the convergence rate, especially at early training, as well as the solution accuracy. To better understand the learning process, we plot the neural network solution in the frequency domain to examine how the network captures successively different frequency bands present in the solution. We consider both forward problems, where the approximate solutions are obtained, as well as inverse problems, where parameters involved in the governing equation are identified. Our simulation results show that the proposed method is a very simple and effective approach to increase the efficiency, robustness and accuracy of the neural network approximation of nonlinear functions as well as solutions of partial differential equations, especially for forward problems. We theoretically prove that in the proposed method, gradient descent algorithms are not attracted to suboptimal critical points or local minima. Furthermore, the proposed adaptive activation functions are shown to accelerate the minimization process of the loss values in standard deep learning benchmarks using CIFAR-10, CIFAR-100, SVHN, MNIST, KMNIST, Fashion-MNIST, and Semeion datasets with and without data augmentation.
Overview
The claims and statistics of this paper haven’t been checked yet.
Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
The claims and statistics of this paper haven’t been checked yet.
Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The self-check needs the paper’s full text.
Citations
Cites
Adam: A Method for Stochastic Optimization
Kingma, D. P., & Ba, J. (2014) · arXiv · doi:10.48550/arxiv.1412.6980 · Cited by 98,946
Gradient-based learning applied to document recognition
Lecun, Y., Bottou, L., Bengio, Y., & Haffner, P. (1998) · Proceedings of the IEEE · doi:10.1109/5.726791 · Cited by 58,805
Raissi, M., Perdikaris, P., & Karniadakis, G. E. (2019) · Journal of Computational Physics · doi:10.1016/j.jcp.2018.10.045 · Cited by 20,384
Self-citationIdentity mappings in deep residual networks
He, K., Zhang, X., Ren, S., & Sun, J. (2016) · Lecture Notes in Computer Science · doi:10.1007/978-3-319-46493-0_38 · Cited by 10,741
Identity Mappings in Deep Residual Networks
He, K., Zhang, X., Ren, S., & Sun, J. (2016) · arXiv · doi:10.48550/arxiv.1603.05027 · Cited by 10,665
Learning long-term dependencies with gradient descent is difficult
Bengio, Y., Simard, P., & Frasconi, P. (1994) · IEEE Transactions on Neural Networks · doi:10.1109/72.279181 · Cited by 8,542
Research works printed but not linked: 19
- Basdevant, 1986, Spectral and finite difference solution of the Burgers equation, Comput. Fluids, 14, 23, 10.1016/0045-7930(86)90036-8 · DOI
- Bateman, 1915, Some recent researches on the motion of fluids, Mon. Weather Rev., 43, 163, 10.1175/1520-0493(1915)43<163:SRROTM>2.0.CO;2 · DOI · Paper page
- Baydin, 2018, Automatic differentiation in machine learning: a survey, J. Mach. Learn. Res., 18, 1
- Berg, 2019, Data-driven discovery of PDEs in complex datasets, J. Comput. Phys., 384, 239, 10.1016/j.jcp.2019.01.036 · DOI
- Burgers, 1948, A mathematical model illustrating the theory of turbulence, 171, 10.1016/S0065-2156(08)70100-5 · DOI · Paper page
- Caudrey, 1975, The sine-Gordon equation as a model classical field theory, Nuovo Cimento, 25, 497, 10.1007/BF02724733 · doi:10.1007/bf02724733
- Lagaris, 1998, Artificial neural network for solving ordinary and partial differential equations, IEEE Trans. Neural Netw., 9, 987, 10.1109/72.712178 · DOI · Paper page
- Li, 2013, The extreme learning machine learning algorithm with tunable activation function, Neural Comput. Appl., 22, 531, 10.1007/s00521-012-0858-9 · DOI
- Owhadi, 2015, Bayesian numerical homogenization, Multiscale Model. Simul., 13, 812, 10.1137/140974596 · DOI
- Parish, 2016, A paradigm for data-driven predictive modeling using field inversion and machine learning, J. Comput. Phys., 305, 758, 10.1016/j.jcp.2015.11.012 · DOI · Paper page
- Qian, 2018, Adaptive activation functions in convolutional neural networks, Neurocomputing, 272, 204, 10.1016/j.neucom.2017.06.070 · DOI · Paper page
- Raissi, 2018, Hidden physics models: machine learning of nonlinear partial differential equations, J. Comput. Phys., 357, 125, 10.1016/j.jcp.2017.11.039 · DOI · Paper page
- Raissi, 2018, Numerical Gaussian processes for time-dependent and nonlinear partial differential equations, SIAM J. Sci. Comput., 40, A172, 10.1137/17M1120762 · DOI · Paper page
- Raissi, 2017, Inferring solutions of differential equations using noisy multi-fidelity data, J. Comput. Phys., 335, 736, 10.1016/j.jcp.2017.01.060 · DOI · Paper page
- Raissi, 2017, Machine learning of linear differential equations using Gaussian processes, J. Comput. Phys., 348, 683, 10.1016/j.jcp.2017.07.050 · DOI · Paper page
- Raissi, 2019, Deep learning of vortex-induced vibrations, J. Fluid Mech., 861, 119, 10.1017/jfm.2018.872 · DOI · Paper page
- Rudy, 2017, Data-driven discovery of partial differential equations, Sci. Adv., 3, 10.1126/sciadv.1602614 · DOI · Paper page
- Shen, 2004, A new multi-output neural model with tunable activation function and its applications, Neural Process. Lett., 20, 85, 10.1007/s11063-004-0637-4 · DOI
- Wazwaz, 2008, New traveling wave solutions to the Boussinesq and the Klein-Gordon equations, Commun. Nonlinear Sci. Numer. Simul., 13, 889, 10.1016/j.cnsns.2006.08.005 · doi:10.1016/j.cnsns.2006.08.005
Cited by
DeepXDE: A Deep Learning Library for Solving Differential Equations
Lu, L., Meng, X., Mao, Z., & Karniadakis, G. E. (2021) · SIAM Review · doi:10.1137/19m1274067 · Cited by 2,512
Self-citationPhysics-informed neural networks (PINNs) for fluid mechanics: a review
Cai, S., Mao, Z., Wang, Z., Yin, M., & Karniadakis, G. E. (2021) · Acta Mechanica Sinica · doi:10.1007/s10409-021-01148-1 · Cited by 2,084
Self-citationPhysics-informed neural networks for high-speed flows
Mao, Z., Jagtap, A. D., & Karniadakis, G. E. (2020) · Computer Methods in Applied Mechanics and Engineering · doi:10.1016/j.cma.2019.112789 · Cited by 1,282
Self-citationActivation functions in deep learning: a comprehensive survey and benchmark
Dubey, S. R., Singh, S. K., & Chaudhuri, B. B. (2022) · Neurocomputing · doi:10.1016/j.neucom.2022.06.111 · Cited by 1,259
Jin, X., Cai, S., Li, H., & Karniadakis, G. E. (2021) · Journal of Computational Physics · doi:10.1016/j.jcp.2020.109951 · Cited by 1,210
Self-citation
Metadata
- Authors
- Jagtap, A. D., Kawaguchi, K., & Karniadakis, G. E.
- Venue
- Journal of Computational Physics
- Publisher
- Elsevier BV
- Published
- 2020-03
- Volume
- 404
- Pages
- 109136
- ISSN
- 0021-9991 (ISSN-L)
- OpenAlex
- W2948551291
- Cited by
- 1,127
- References
- 6
- FWCI
- 25.44
- Type
- journal article
- Language
- en
- Open access
- bronze
- Licence
- www.elsevier.com/tdm/userlicense/1.0/ (tdm)
- www.elsevier.com/open-access/userlicense/1.0/ (am, 314 d)
- Funders
- U.S. Department of Energy
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
How to cite
Cite CoChat for the checks on this page, and the paper for its own findings.
- This page
- CoChat. (2026). Citation check: Adaptive activation functions accelerate convergence in deep and physics-informed neural networks. https://app.cochat.ai/papers/10.1016/j.jcp.2019.109136
- The paper
- Jagtap, A. D., Kawaguchi, K., & Karniadakis, G. E. (2020). Adaptive activation functions accelerate convergence in deep and physics-informed neural networks. Journal of Computational Physics. https://doi.org/10.1016/j.jcp.2019.109136
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.