{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,7,17]],"date-time":"2026-07-17T18:02:52Z","timestamp":1784311372795,"version":"3.55.0"},"reference-count":45,"publisher":"Elsevier BV","license":[{"start":{"date-parts":[[2026,11,1]],"date-time":"2026-11-01T00:00:00Z","timestamp":1793491200000},"content-version":"tdm","delay-in-days":0,"URL":"https:\/\/www.elsevier.com\/tdm\/userlicense\/1.0\/"},{"start":{"date-parts":[[2026,11,1]],"date-time":"2026-11-01T00:00:00Z","timestamp":1793491200000},"content-version":"tdm","delay-in-days":0,"URL":"https:\/\/www.elsevier.com\/legal\/tdmrep-license"},{"start":{"date-parts":[[2026,6,16]],"date-time":"2026-06-16T00:00:00Z","timestamp":1781568000000},"content-version":"vor","delay-in-days":0,"URL":"http:\/\/creativecommons.org\/licenses\/by-nc-nd\/4.0\/"}],"funder":[{"DOI":"10.13039\/501100014434","name":"National Agency for Academic Exchange","doi-asserted-by":"publisher","id":[{"id":"10.13039\/501100014434","id-type":"DOI","asserted-by":"publisher"}]}],"content-domain":{"domain":["elsevier.com","sciencedirect.com"],"crossmark-restriction":true},"short-container-title":["Information Sciences"],"published-print":{"date-parts":[[2026,11]]},"DOI":"10.1016\/j.ins.2026.123796","type":"journal-article","created":{"date-parts":[[2026,6,16]],"date-time":"2026-06-16T23:11:28Z","timestamp":1781651488000},"page":"123796","update-policy":"https:\/\/doi.org\/10.1016\/elsevier_cm_policy","source":"Crossref","is-referenced-by-count":0,"special_numbering":"C","title":["Fully tensorial approach to hypercomplex-valued neural networks"],"prefix":"10.1016","volume":"755","author":[{"ORCID":"https:\/\/orcid.org\/0000-0002-4370-3326","authenticated-orcid":false,"given":"Agnieszka","family":"Niemczynowicz","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-6390-4627","authenticated-orcid":false,"given":"Rados\u0142aw Antoni","family":"Kycia","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"78","reference":[{"issue":"1\u20132","key":"10.1016\/j.ins.2026.123796_bib0005","doi-asserted-by":"crossref","first-page":"33","DOI":"10.1016\/0166-218X(95)00075-3","article-title":"Representation theory and invariant neural networks","volume":"69","author":"Wood","year":"1996","journal-title":"Discrete Appl. Math."},{"key":"10.1016\/j.ins.2026.123796_bib0010","doi-asserted-by":"crossref","first-page":"30545","DOI":"10.1007\/s11042-022-12683-9","article-title":"Geometric machine learning: research and applications","volume":"81","author":"Cao","year":"2022","journal-title":"Multimed. Tools Appl."},{"key":"10.1016\/j.ins.2026.123796_bib0015","doi-asserted-by":"crossref","first-page":"35929","DOI":"10.1109\/ACCESS.2020.2975067","article-title":"A comprehensive survey on geometric deep learning","volume":"8","author":"Cao","year":"2020","journal-title":"IEEE Access"},{"issue":"2","key":"10.1016\/j.ins.2026.123796_bib0020","doi-asserted-by":"crossref","first-page":"255","DOI":"10.1090\/S0273-0979-09-01249-X","article-title":"Topology and data","volume":"46","author":"Carlsson","year":"2009","journal-title":"Bull. Amer. Math. Soc."},{"key":"10.1016\/j.ins.2026.123796_bib0025","series-title":"Proceedings of the 12th USENIX Conference on Operating Systems Design and Implementation (OSDI\u201916)","first-page":"265","article-title":"TensorFlow: a system for large-scale machine learning","author":"Abadi","year":"2016"},{"key":"10.1016\/j.ins.2026.123796_bib0030","series-title":"Proceedings of the 33rd International Conference on Neural Information Processing Systems","first-page":"8026","article-title":"PyTorch: an imperative style, high-performance deep learning library","volume":"vol. 721","author":"Paszke","year":"2019"},{"issue":"26","key":"10.1016\/j.ins.2026.123796_bib0035","first-page":"1","article-title":"TensorLy: tensor learning in Python","volume":"20","author":"Kossaifi","year":"2019","journal-title":"J. Mach. Learn. Res."},{"key":"10.1016\/j.ins.2026.123796_bib0040","doi-asserted-by":"crossref","first-page":"2017","DOI":"10.1007\/s10586-021-03240-4","article-title":"DLBench: a comprehensive experimental evaluation of deep learning frameworks","volume":"24","author":"Elshawi","year":"2021","journal-title":"Cluster Comput"},{"key":"10.1016\/j.ins.2026.123796_bib0045","doi-asserted-by":"crossref","first-page":"3351","DOI":"10.1109\/TSP.2021.3084537","article-title":"Algebraic neural networks: stability to deformations","volume":"69","author":"Parada-Mayorga","year":"2021","journal-title":"IEEE Trans. Signal Process."},{"issue":"7","key":"10.1016\/j.ins.2026.123796_bib0050","doi-asserted-by":"crossref","first-page":"925","DOI":"10.1016\/j.neunet.2008.03.004","article-title":"On clifford neurons and clifford multi-layer perceptrons","volume":"21","author":"Buchholz","year":"2008","journal-title":"Neural Netw."},{"key":"10.1016\/j.ins.2026.123796_bib0055","series-title":"Proceedings of the 40th International Conference on Machine Learning (ICML\u201923), Vol. 202 JMLR.org, Article 1219","first-page":"29306","article-title":"Geometric clifford algebra networks","author":"Ruhe","year":"2023"},{"key":"10.1016\/j.ins.2026.123796_bib0060","series-title":"2018 International Joint Conference on Neural Networks (IJCNN), Rio de Janeiro","first-page":"1","article-title":"Deep quaternion networks","author":"Gaudet","year":"2018"},{"key":"10.1016\/j.ins.2026.123796_bib0065","series-title":"2020 International Joint Conference on Neural Networks (IJCNN)","first-page":"1","article-title":"Constructing convolutional neural networks based on quaternion","author":"Hongo","year":"2020"},{"key":"10.1016\/j.ins.2026.123796_bib0070","first-page":"1","article-title":"PHNNs: lightweight neural networks via parameterized hypercomplex convolutions","author":"Grassucci","year":"2022","journal-title":"IEEE Trans. Neural Netw. Learn. Syst."},{"key":"10.1016\/j.ins.2026.123796_bib0075","series-title":"9th International Conference on Learning Representations, ICLR 2021","article-title":"Beyond fully-connected layers with quaternions: parameterization of hypercomplex multiplications with1\/nparameters","author":"Zhang","year":"2021"},{"key":"10.1016\/j.ins.2026.123796_bib0080","series-title":"Neural Networks in Multidimensional Domains, Lecture Notes in Control and Information Sciences","author":"Arena","year":"1998"},{"key":"10.1016\/j.ins.2026.123796_bib0085","series-title":"Advanced Computational Applications of Geometric Algebra. ICACGA 2022. Lecture Notes in Computer Science","article-title":"Clifford convolutional neural networks for lymphoblast image classification","volume":"vol. 13771","author":"Vieira","year":"2024"},{"key":"10.1016\/j.ins.2026.123796_bib0090","doi-asserted-by":"crossref","DOI":"10.1038\/s41598-025-08957-5","article-title":"4D hypercomplex-valued neural network in multivariate time series forecasting","volume":"15","author":"Kycia","year":"2025","journal-title":"Sci. Rep."},{"key":"10.1016\/j.ins.2026.123796_bib0095","doi-asserted-by":"crossref","first-page":"1042","DOI":"10.1002\/mma.2660","article-title":"Algebraic foundations of split hypercomplex nonlinear adaptive filtering","volume":"36","author":"Hitzer","year":"2013","journal-title":"Math. Meth. Appl. Sci."},{"key":"10.1016\/j.ins.2026.123796_bib0100","series-title":"Proceedings of the Twenty-Eighth International Joint Conference on Artificial Intelligence Main Track","first-page":"4313","article-title":"Quaternion collaborative filtering for recommendation","author":"Zhang","year":"2019"},{"key":"10.1016\/j.ins.2026.123796_bib0105","author":"Chen"},{"key":"10.1016\/j.ins.2026.123796_bib0110","series-title":"Artificial Neural Networks and Machine Learning \u2013 ICANN 2021: 30th International Conference on Artificial Neural Networks, (14-17 September 2021), Proceedings, Part III","first-page":"204","article-title":"Parameterized hypercomplex graph neural networks for graph classification","author":"Le","year":"2021"},{"issue":"2","key":"10.1016\/j.ins.2026.123796_bib0115","doi-asserted-by":"crossref","DOI":"10.1109\/72.914526","article-title":"A comparison between HMLP and HRBF for attitude control","volume":"12","author":"Fortuna","year":"2001","journal-title":"IEEE Trans. Neural Netw."},{"key":"10.1016\/j.ins.2026.123796_bib0120","series-title":"Proceedings of the 28th International Conference on Computational Linguistics","first-page":"530","article-title":"Knowledge graph embeddings in geometric algebras International committee on computational linguistics","author":"Xu","year":"2020"},{"key":"10.1016\/j.ins.2026.123796_bib0125","first-page":"2731","article-title":"Quaternion knowledge graph embeddings","volume":"32","author":"Zhang","year":"2019","journal-title":"Adv. Neural Inf. Process. Syst."},{"key":"10.1016\/j.ins.2026.123796_bib0130","series-title":"5th Workshop on Knowledge Graphs and Big Data in IEEE BigData","article-title":"HyperComplEx: adaptive multi-space knowledge graph embeddings","author":"Gajjar","year":"2025"},{"key":"10.1016\/j.ins.2026.123796_bib0135","doi-asserted-by":"crossref","first-page":"3221","DOI":"10.3390\/app14083221","article-title":"RiQ-KGC: relation instantiation enhanced quaternionic attention for complex-relation knowledge graph completion","volume":"14","author":"Wang","year":"2024","journal-title":"Appl. Sci."},{"key":"10.1016\/j.ins.2026.123796_bib0140","doi-asserted-by":"crossref","first-page":"66526","DOI":"10.1109\/ACCESS.2026.3687751","article-title":"Multi-view hypercomplex learning for breast cancer screening","volume":"14","author":"Lopez","year":"2026","journal-title":"IEEE Access"},{"key":"10.1016\/j.ins.2026.123796_bib0145","series-title":"Proc. Interspeech 2018","first-page":"22","article-title":"Quaternion convolutional neural networks for end-to-end automatic speech recognition","author":"Parcollet","year":"2018"},{"key":"10.1016\/j.ins.2026.123796_bib0150","doi-asserted-by":"crossref","DOI":"10.1016\/j.sigpro.2024.109504","article-title":"Quaternion matrix completion using untrained quaternion convolutional neural network for color image inpainting","volume":"221","author":"Miao","year":"2024","journal-title":"Signal Process."},{"key":"10.1016\/j.ins.2026.123796_bib0155","doi-asserted-by":"crossref","first-page":"111","DOI":"10.1016\/j.neucom.2020.12.034","article-title":"Hypercomplex-valued recurrent correlation neural networks","volume":"432","author":"Valle","year":"2021","journal-title":"Neurocomputing"},{"key":"10.1016\/j.ins.2026.123796_bib0160","series-title":"Computer Vision - ECCV 2018. ECCV 2018. Lecture Notes in Computer Science","article-title":"Quaternion convolutional neural networks","volume":"vol. 11212","author":"Zhu","year":"2018"},{"key":"10.1016\/j.ins.2026.123796_bib0165","series-title":"Algebra","author":"Aluffi","year":"2009"},{"key":"10.1016\/j.ins.2026.123796_bib0170","doi-asserted-by":"crossref","DOI":"10.1016\/j.softx.2025.102163","article-title":"KHNN: hypercomplex-valued neural networks computations via keras using TensorFlow and PyTorch","volume":"30","author":"Niemczynowicz","year":"2025","journal-title":"Softwarex"},{"key":"10.1016\/j.ins.2026.123796_bib0175","series-title":"Spinors and Space-Time, Volume 1: Two-Spinor Calculus and Relativistic Fields","author":"Penrose","year":"1984"},{"key":"10.1016\/j.ins.2026.123796_bib0180","series-title":"Representation Theory. a First Course","author":"Fulton","year":"1991"},{"key":"10.1016\/j.ins.2026.123796_bib0185","unstructured":"Hypercomplex Keras: https:\/\/pypi.org\/project\/HypercomplexKeras\/ (accessed 9 September 2024)."},{"key":"10.1016\/j.ins.2026.123796_bib0190","series-title":"Proceedings of the Thirteenth International Conference on Artificial Intelligence and Statistics, PMLR","first-page":"249","article-title":"Understanding the difficulty of training deep feedforward neural networks","volume":"vol. 9","author":"Glorot","year":"2010"},{"key":"10.1016\/j.ins.2026.123796_bib0195","series-title":"2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), Las Vegas","first-page":"4013","article-title":"Fast algorithms for convolutional neural networks","author":"Lavin","year":"2016"},{"key":"10.1016\/j.ins.2026.123796_bib0200","doi-asserted-by":"crossref","DOI":"10.1016\/j.neunet.2024.106632","article-title":"Universal approximation theorem for vector- and hypercomplex-valued neural networks","volume":"180","author":"Valle","year":"2024","journal-title":"Neural Netw."},{"key":"10.1016\/j.ins.2026.123796_bib0205","series-title":"Functional Analysis","author":"Bachman","year":"1998"},{"issue":"4","key":"10.1016\/j.ins.2026.123796_bib0210","doi-asserted-by":"crossref","first-page":"303","DOI":"10.1007\/BF02551274","article-title":"Approximation by superpositions of a sigmoidal function","volume":"2","author":"Cybenko","year":"1989","journal-title":"Math. Control Signals Syst."},{"issue":"3","key":"10.1016\/j.ins.2026.123796_bib0215","doi-asserted-by":"crossref","first-page":"930","DOI":"10.1109\/18.256500","article-title":"Universal approximation bounds for superpositions of a sigmoidal function","volume":"39","author":"Barron","year":"1993","journal-title":"IEEE Trans. Inf. theory"},{"issue":"2","key":"10.1016\/j.ins.2026.123796_bib0220","doi-asserted-by":"crossref","first-page":"251","DOI":"10.1016\/0893-6080(91)90009-T","article-title":"Approximation capabilities of multilayer feedforward networks","volume":"4","author":"Hornik","year":"1991","journal-title":"Neural Netw."},{"issue":"6","key":"10.1016\/j.ins.2026.123796_bib0225","doi-asserted-by":"crossref","first-page":"861","DOI":"10.1016\/S0893-6080(05)80131-5","article-title":"Multilayer feedforward networks with a nonpolynomial activation function can approximate any function","volume":"6","author":"Leshno","year":"1993","journal-title":"Neural Netw."}],"container-title":["Information Sciences"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/api.elsevier.com\/content\/article\/PII:S0020025526007279?httpAccept=text\/xml","content-type":"text\/xml","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/api.elsevier.com\/content\/article\/PII:S0020025526007279?httpAccept=text\/plain","content-type":"text\/plain","content-version":"vor","intended-application":"text-mining"}],"deposited":{"date-parts":[[2026,7,17]],"date-time":"2026-07-17T17:04:40Z","timestamp":1784307880000},"score":1,"resource":{"primary":{"URL":"https:\/\/linkinghub.elsevier.com\/retrieve\/pii\/S0020025526007279"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,11]]},"references-count":45,"alternative-id":["S0020025526007279"],"URL":"https:\/\/doi.org\/10.1016\/j.ins.2026.123796","relation":{},"ISSN":["0020-0255"],"issn-type":[{"value":"0020-0255","type":"print"}],"subject":[],"published":{"date-parts":[[2026,11]]},"assertion":[{"value":"Elsevier","name":"publisher","label":"This article is maintained by"},{"value":"Fully tensorial approach to hypercomplex-valued neural networks","name":"articletitle","label":"Article Title"},{"value":"Information Sciences","name":"journaltitle","label":"Journal Title"},{"value":"https:\/\/doi.org\/10.1016\/j.ins.2026.123796","name":"articlelink","label":"CrossRef DOI link to publisher maintained version"},{"value":"article","name":"content_type","label":"Content Type"},{"value":"\u00a9 2026 The Author(s). Published by Elsevier Inc.","name":"copyright","label":"Copyright"}],"article-number":"123796"}}