SciANN: A Keras/TensorFlow wrapper for scientific computations and physics-informed deep learning using artificial neural networks
© 2020 Elsevier B.V. In this paper, we introduce SciANN, a Python package for scientific computing and physics-informed deep learning using artificial neural networks. SciANN uses the widely used deep-learning packages TensorFlow and Keras to build deep neural networks and optimization models, thus...
| Main Authors: | , |
|---|---|
| Other Authors: | |
| Format: | Article |
| Language: | English |
| Published: |
Elsevier BV
2021
|
| Online Access: | https://hdl.handle.net/1721.1/133000 |
| _version_ | 1826216165243879424 |
|---|---|
| author | Haghighat, Ehsan Juanes, Ruben |
| author2 | Massachusetts Institute of Technology. Department of Civil and Environmental Engineering |
| author_facet | Massachusetts Institute of Technology. Department of Civil and Environmental Engineering Haghighat, Ehsan Juanes, Ruben |
| author_sort | Haghighat, Ehsan |
| collection | MIT |
| description | © 2020 Elsevier B.V. In this paper, we introduce SciANN, a Python package for scientific computing and physics-informed deep learning using artificial neural networks. SciANN uses the widely used deep-learning packages TensorFlow and Keras to build deep neural networks and optimization models, thus inheriting many of Keras's functionalities, such as batch optimization and model reuse for transfer learning. SciANN is designed to abstract neural network construction for scientific computations and solution and discovery of partial differential equations (PDE) using the physics-informed neural networks (PINN) architecture, therefore providing the flexibility to set up complex functional forms. We illustrate, in a series of examples, how the framework can be used for curve fitting on discrete data, and for solution and discovery of PDEs in strong and weak forms. We summarize the features currently available in SciANN, and also outline ongoing and future developments. |
| first_indexed | 2024-09-23T16:42:59Z |
| format | Article |
| id | mit-1721.1/133000 |
| institution | Massachusetts Institute of Technology |
| language | English |
| last_indexed | 2024-09-23T16:42:59Z |
| publishDate | 2021 |
| publisher | Elsevier BV |
| record_format | dspace |
| spelling | mit-1721.1/1330002024-06-06T00:02:34Z SciANN: A Keras/TensorFlow wrapper for scientific computations and physics-informed deep learning using artificial neural networks Haghighat, Ehsan Juanes, Ruben Massachusetts Institute of Technology. Department of Civil and Environmental Engineering © 2020 Elsevier B.V. In this paper, we introduce SciANN, a Python package for scientific computing and physics-informed deep learning using artificial neural networks. SciANN uses the widely used deep-learning packages TensorFlow and Keras to build deep neural networks and optimization models, thus inheriting many of Keras's functionalities, such as batch optimization and model reuse for transfer learning. SciANN is designed to abstract neural network construction for scientific computations and solution and discovery of partial differential equations (PDE) using the physics-informed neural networks (PINN) architecture, therefore providing the flexibility to set up complex functional forms. We illustrate, in a series of examples, how the framework can be used for curve fitting on discrete data, and for solution and discovery of PDEs in strong and weak forms. We summarize the features currently available in SciANN, and also outline ongoing and future developments. 2021-10-15T18:17:56Z 2021-10-15T18:17:56Z 2020-11 2020-10 2021-10-15T17:01:06Z Article http://purl.org/eprint/type/JournalArticle 0045-7825 https://hdl.handle.net/1721.1/133000 Ehsan Haghighat, Ruben Juanes, SciANN: A Keras/TensorFlow wrapper for scientific computations and physics-informed deep learning using artificial neural networks, Computer Methods in Applied Mechanics and Engineering, Volume 373, 2021 en 10.1016/J.CMA.2020.113552 Computer Methods in Applied Mechanics and Engineering Creative Commons Attribution-NonCommercial-NoDerivs License http://creativecommons.org/licenses/by-nc-nd/4.0/ application/pdf Elsevier BV arXiv |
| spellingShingle | Haghighat, Ehsan Juanes, Ruben SciANN: A Keras/TensorFlow wrapper for scientific computations and physics-informed deep learning using artificial neural networks |
| title | SciANN: A Keras/TensorFlow wrapper for scientific computations and physics-informed deep learning using artificial neural networks |
| title_full | SciANN: A Keras/TensorFlow wrapper for scientific computations and physics-informed deep learning using artificial neural networks |
| title_fullStr | SciANN: A Keras/TensorFlow wrapper for scientific computations and physics-informed deep learning using artificial neural networks |
| title_full_unstemmed | SciANN: A Keras/TensorFlow wrapper for scientific computations and physics-informed deep learning using artificial neural networks |
| title_short | SciANN: A Keras/TensorFlow wrapper for scientific computations and physics-informed deep learning using artificial neural networks |
| title_sort | sciann a keras tensorflow wrapper for scientific computations and physics informed deep learning using artificial neural networks |
| url | https://hdl.handle.net/1721.1/133000 |
| work_keys_str_mv | AT haghighatehsan sciannakerastensorflowwrapperforscientificcomputationsandphysicsinformeddeeplearningusingartificialneuralnetworks AT juanesruben sciannakerastensorflowwrapperforscientificcomputationsandphysicsinformeddeeplearningusingartificialneuralnetworks |