{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,6,15]],"date-time":"2026-06-15T23:04:26Z","timestamp":1781564666656,"version":"3.54.5"},"reference-count":30,"publisher":"Springer Science and Business Media LLC","issue":"5","license":[{"start":{"date-parts":[[2024,3,26]],"date-time":"2024-03-26T00:00:00Z","timestamp":1711411200000},"content-version":"tdm","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0"},{"start":{"date-parts":[[2024,3,26]],"date-time":"2024-03-26T00:00:00Z","timestamp":1711411200000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0"}],"funder":[{"DOI":"10.13039\/501100002551","name":"Seoul National University","doi-asserted-by":"crossref","id":[{"id":"10.13039\/501100002551","id-type":"DOI","asserted-by":"crossref"}]}],"content-domain":{"domain":["link.springer.com"],"crossmark-restriction":false},"short-container-title":["Mach Learn"],"published-print":{"date-parts":[[2024,5]]},"abstract":"<jats:title>Abstract<\/jats:title><jats:p>Recently, several types of neural operators have been developed, including deep operator networks, graph neural operators, and Multiwavelet-based operators. Compared with these models, the Fourier neural operator (FNO), a physics-inspired machine learning method, is computationally efficient and can learn nonlinear operators between function spaces independent of a certain finite basis. This study investigated the bounding of the Rademacher complexity of the FNO based on specific group norms. Using capacity based on these norms, we bound the generalization error of the model. In addition, we investigate the correlation between the empirical generalization error and the proposed capacity of FNO. We infer that the type of group norm determines the information about the weights and architecture of the FNO model stored in capacity. The experimental results offer insight into the impact of the number of modes used in the FNO model on the generalization error. The results confirm that our capacity is an effective index for estimating generalization errors.<\/jats:p>","DOI":"10.1007\/s10994-024-06533-y","type":"journal-article","created":{"date-parts":[[2024,3,26]],"date-time":"2024-03-26T17:03:29Z","timestamp":1711472609000},"page":"2467-2498","update-policy":"https:\/\/doi.org\/10.1007\/springer_crossmark_policy","source":"Crossref","is-referenced-by-count":8,"title":["Bounding the Rademacher complexity of Fourier neural operators"],"prefix":"10.1007","volume":"113","author":[{"given":"Taeyoung","family":"Kim","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Myungjoo","family":"Kang","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"297","published-online":{"date-parts":[[2024,3,26]]},"reference":[{"key":"6533_CR1","unstructured":"Awasthi, P., Frank, N., & Mohri, M. (2020). On the rademacher complexity of linear hypothesis sets. arXiv arXiv:2007.11045"},{"key":"6533_CR2","unstructured":"Bartlett, P. L., Foster, D. J. & Telgarsky, M. (2021). Spectrally-normalized margin bounds for neural networks. Advances in Neural Information Processing Systems"},{"key":"6533_CR3","doi-asserted-by":"publisher","first-page":"686","DOI":"10.1016\/j.jcp.2021.110514","volume":"443","author":"Z Cai","year":"2021","unstructured":"Cai, Z., Chen, J., & Liu, M. (2021). Least-squares ReLU neural network (LSNN) method for linear advection-reaction equation. Journal of Computational Physics, 443, 686\u2013707.","journal-title":"Journal of Computational Physics"},{"key":"6533_CR4","unstructured":"Gopalani, P., Karmakar, S. & Mukherjee, A. (2022). Capacity bounds for the deeponet method of solving differential equations. arXiv arXiv:2205.11359"},{"key":"6533_CR5","first-page":"24048","volume":"34","author":"G Gupta","year":"2021","unstructured":"Gupta, G., Xiao, X., & Bogdan, P. (2021). Multiwavelet-based operator learning for differential equations. Advances in Neural Information Processing Systems, 34, 24048\u201324062.","journal-title":"Advances in Neural Information Processing Systems"},{"key":"6533_CR6","unstructured":"Hao, C., Zhanfeng, M., Zhouwang, Y. & Xiao, W. (2019). Theoretical investigation of generalization bound for residual networks. In International joint conferences on artificial intelligence organization, pp. 2081\u20132087"},{"key":"6533_CR7","doi-asserted-by":"crossref","unstructured":"Jakubovitz, D., Giryes, R., Rodrigues, M. R. D. (2019). Generalization Error in Deep Learning. Birkhuser Cham, GEWERBESTRASSE 11, Cham, Switzerland.","DOI":"10.1007\/978-3-319-73074-5_5"},{"key":"6533_CR8","unstructured":"Kovachki, N., Li, Z., Liu, B., Azizzadenesheli, K., Bhattacharya, K., Stuart, A. & Anandkumar, A. (2021). Neural operator: Learning maps between function spaces. arXiv arXiv:2108.08481"},{"issue":"290","key":"6533_CR9","first-page":"1","volume":"22","author":"N Kovachki","year":"2021","unstructured":"Kovachki, N., Lanthaler, S., & Mishra, S. (2021). On universal approximation and error bounds for Fourier neural operators. Journal of Machine Learning Research, 22(290), 1\u201376.","journal-title":"Journal of Machine Learning Research"},{"issue":"5","key":"6533_CR10","doi-asserted-by":"publisher","first-page":"2995","DOI":"10.1109\/TIT.2019.2893916","volume":"65","author":"Y Lei","year":"2019","unstructured":"Lei, Y., Dogan, U., Zhou, D., & Kloft, M. (2019). Data-dependent generalization bounds for multi-class classification. IEEE Transactions on Information Theory, 65(5), 2995\u20133021.","journal-title":"IEEE Transactions on Information Theory"},{"key":"6533_CR11","unstructured":"Li, Z., Kovachki, N., Azizzadenesheli, K., Liu, B., Bhattacharya, K., Stuart, A. & Anandkumar, A. (2020). Neural operator: Graph kernel network for partial differential equations. ICLR 2020 Workshop ODE\/PDE+DL"},{"key":"6533_CR12","unstructured":"Li, Z., Kovachki, N., Azizzadenesheli, K., Liu, B., Bhattacharya, K., Stuart, A. & Anandkumar, A. (2021) Fourier neural operator for parametric partial differential equations. ICLR 2021."},{"key":"6533_CR13","first-page":"888","volume":"89","author":"T Liang","year":"2019","unstructured":"Liang, T., Poggio, T., Rakhlin, A., & Stokes, J. (2019). Fisher-rao metric, geometry, and complexity of neural networks. Proceedings of the Twenty-Second International Conference on Artificial Intelligence and Statistics, 89, 888\u2013896.","journal-title":"Proceedings of the Twenty-Second International Conference on Artificial Intelligence and Statistics"},{"key":"6533_CR14","unstructured":"Long, P. M. & Sedghi, H. (2020). Generalization bounds for deep convolutional neural networks. ICLR 2020."},{"issue":"3","key":"6533_CR15","doi-asserted-by":"publisher","first-page":"218","DOI":"10.1038\/s42256-021-00302-5","volume":"3","author":"L Lu","year":"2021","unstructured":"Lu, L., Jin, P., Pang, G., Zhang, Z., & Karniadakis, G. E. (2021). Learning nonlinear operators via deeponet based on the universal approximation theorem of operators. Nature Machine Intelligence, 3(3), 218\u2013229.","journal-title":"Nature Machine Intelligence"},{"key":"6533_CR16","unstructured":"Lv, S. (2021). Generalization bounds for graph convolutional neural networks via rademacher complexity. arXiv arXiv:2102.10234"},{"key":"6533_CR17","doi-asserted-by":"crossref","unstructured":"Maurer, A. (2016). A vector-contraction inequality for rademacher complexities. arXiv arXiv:1605.00251","DOI":"10.1007\/978-3-319-46379-7_1"},{"key":"6533_CR18","first-page":"1233","volume":"108","author":"C Minshuo","year":"2020","unstructured":"Minshuo, C., Xingguo, L., & Tuo, Z. (2020). On generalization bounds of a family of recurrent neural networks. Proceedings of Machine Learning Research, 108, 1233\u20131243.","journal-title":"Proceedings of Machine Learning Research"},{"key":"6533_CR19","first-page":"1376","volume":"40","author":"B Neyshabur","year":"2015","unstructured":"Neyshabur, B., Tomioka, R., & Srebro, N. (2015). Norm-based capacity control in neural networks. Proceedings of Machine Learning Research, 40, 1376\u20131401.","journal-title":"Proceedings of Machine Learning Research"},{"key":"6533_CR20","unstructured":"Pathak, J., subramanian, S., Harrington, P., Raja, S., Chattopadhyay, A., Mardani, M., Kurth, T., Hall, D., Li, Z., Azizzadenesheli, K., Hassanzadeh, P., Kashinath, K. & Anandkumar, A. (2022). Fourcastnet: A global data-driven high-resolution weather model using adaptive fourier neural operators. arXiv arXiv:2202.11214"},{"key":"6533_CR21","first-page":"18420","volume":"34","author":"H Petzka","year":"2021","unstructured":"Petzka, H., Kamp, M., Adilova, L., Sminchisescu, C., & Boley, M. (2021). Relative flatness and generalization. Advances in Neural Information Processing Systems, 34, 18420\u201318432.","journal-title":"Advances in Neural Information Processing Systems"},{"key":"6533_CR22","doi-asserted-by":"publisher","first-page":"686","DOI":"10.1016\/j.jcp.2018.10.045","volume":"378","author":"M Raissi","year":"2019","unstructured":"Raissi, M., Perdikaris, P., & Karniadakis, G. E. (2019). Physics-informed neural networks: A deep learning framework for solving forward and inverse problems involving nonlinear partial differential equations. Journal of Computational Physics, 378, 686\u2013707.","journal-title":"Journal of Computational Physics"},{"key":"6533_CR23","first-page":"847","volume":"145","author":"M Seleznova","year":"2021","unstructured":"Seleznova, M., & Kutyniok, G. (2021). Analyzing finite neural networks: Can we trust neural tangent kernel theory? Proceedings of Machine Learning Research, 145, 847\u2013867.","journal-title":"Proceedings of Machine Learning Research"},{"key":"6533_CR24","doi-asserted-by":"crossref","unstructured":"Shalev-Shwartz, S., & Ben-David, S. (2014). Understanding Machine Learning: From Theory to Algorithms. Cambridge University Press.","DOI":"10.1017\/CBO9781107298019"},{"key":"6533_CR25","first-page":"69","volume":"168","author":"ED Sontag","year":"1998","unstructured":"Sontag, E. D. (1998). Vc dimension of neural networks. NATO ASI Series F Computer and Systems Sciences, 168, 69\u201396.","journal-title":"NATO ASI Series F Computer and Systems Sciences"},{"issue":"11","key":"6533_CR26","doi-asserted-by":"publisher","first-page":"1134","DOI":"10.1145\/1968.1972","volume":"27","author":"LG Valiant","year":"1984","unstructured":"Valiant, L. G. (1984). A theory of the learnable. Communications of the ACM, 27(11), 1134\u20131142.","journal-title":"Communications of the ACM"},{"issue":"5","key":"6533_CR27","doi-asserted-by":"publisher","first-page":"988","DOI":"10.1109\/72.788640","volume":"10","author":"VN Vapnik","year":"1999","unstructured":"Vapnik, V. N. (1999). An overview of statistical learning theory. IEEE Transactions on Neural Networks, 10(5), 988\u2013999.","journal-title":"IEEE Transactions on Neural Networks"},{"issue":"6","key":"6533_CR28","doi-asserted-by":"publisher","first-page":"1755","DOI":"10.4310\/CMS.2020.v18.n6.a10","volume":"18","author":"E Weinan","year":"2020","unstructured":"Weinan, E., Chao, M., & Qingcan, W. (2020). Rademacher complexity and the generalization error of residual networks. Communications in Mathematical Sciences, 18(6), 1755\u20131774.","journal-title":"Communications in Mathematical Sciences"},{"issue":"1","key":"6533_CR29","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1007\/s40304-018-0127-z","volume":"6","author":"E Weinan","year":"2018","unstructured":"Weinan, E., & Yu, B. (2018). The deep ritz method: A deep learning-based numerical algorithm for solving variational problems. Communications in Mathematics and Statistics, 6(1), 1\u201312.","journal-title":"Communications in Mathematics and Statistics"},{"key":"6533_CR30","doi-asserted-by":"publisher","DOI":"10.1016\/j.advwatres.2022.104180","volume":"163","author":"G Wen","year":"2022","unstructured":"Wen, G., Li, Z., Azizzadenesheli, K., Anandkumar, A., & Benson, S. M. (2022). U-fno-an enhanced Fourier neural operator-based deep-learning model for multiphase flow. Advances in Water Resources, 163, 104180.","journal-title":"Advances in Water Resources"}],"container-title":["Machine Learning"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/link.springer.com\/content\/pdf\/10.1007\/s10994-024-06533-y.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/link.springer.com\/article\/10.1007\/s10994-024-06533-y\/fulltext.html","content-type":"text\/html","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/link.springer.com\/content\/pdf\/10.1007\/s10994-024-06533-y.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2024,5,2]],"date-time":"2024-05-02T18:20:42Z","timestamp":1714674042000},"score":1,"resource":{"primary":{"URL":"https:\/\/link.springer.com\/10.1007\/s10994-024-06533-y"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2024,3,26]]},"references-count":30,"journal-issue":{"issue":"5","published-print":{"date-parts":[[2024,5]]}},"alternative-id":["6533"],"URL":"https:\/\/doi.org\/10.1007\/s10994-024-06533-y","relation":{},"ISSN":["0885-6125","1573-0565"],"issn-type":[{"value":"0885-6125","type":"print"},{"value":"1573-0565","type":"electronic"}],"subject":[],"published":{"date-parts":[[2024,3,26]]},"assertion":[{"value":"3 October 2022","order":1,"name":"received","label":"Received","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"25 December 2023","order":2,"name":"revised","label":"Revised","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"16 February 2024","order":3,"name":"accepted","label":"Accepted","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"26 March 2024","order":4,"name":"first_online","label":"First Online","group":{"name":"ArticleHistory","label":"Article History"}},{"order":1,"name":"Ethics","group":{"name":"EthicsHeading","label":"Declarations"}},{"value":"The authors declare that they have no conflict of interest.","order":2,"name":"Ethics","group":{"name":"EthicsHeading","label":"Conflict of interest"}},{"value":"Not applicable.","order":3,"name":"Ethics","group":{"name":"EthicsHeading","label":"Ethical approval"}},{"value":"Not applicable.","order":4,"name":"Ethics","group":{"name":"EthicsHeading","label":"Consent to participate"}},{"value":"Not applicable.","order":5,"name":"Ethics","group":{"name":"EthicsHeading","label":"Consent for publication"}}]}}