{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2025,10,22]],"date-time":"2025-10-22T23:36:57Z","timestamp":1761176217312,"version":"build-2065373602"},"reference-count":0,"publisher":"IOS Press","isbn-type":[{"value":"9781643686318","type":"electronic"}],"license":[{"start":{"date-parts":[[2025,10,21]],"date-time":"2025-10-21T00:00:00Z","timestamp":1761004800000},"content-version":"unspecified","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by-nc\/4.0\/"}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":[],"published-print":{"date-parts":[[2025,10,21]]},"abstract":"<jats:p>This study aims to develop identification schemes for inducing an attention mechanism into the modeling of multi-input multi-output nonlinear processes. In particular, the problem under consideration is to improve the performance of the neural network by focusing on relevant information on the dynamics of the identified system. The idea is to apply attention to selective assignment of significance for parts of input data to intensify the participation of the most informative system observations in the training of the neural model. To achieve this goal, two dedicated attention schemes are properly tailored to the regression task and separately investigated. First, with the attentiveness, representing how the input data affects the system state, applied to the model input, and the second one, with the attention used to correct the modeled system state. Both proposed solutions can be relatively easily embedded into the structure of the original state-space neural network used for system identification. This provides additional flexibility of model design, leading to a significant improvement of model generalization properties, especially required in practical situations when we cope with multi-output systems with strong interactions or a high level of noise. Finally, the approach is verified on the relevant real-world examples of modeling complex physical processes such as control of an intelligent valve, deflection of a cantilever beam, or spark-ignition engine. This provides a proper insight into the adaptation of the concept of attention towards the modeling and identification in the framework of industrial applications.<\/jats:p>","DOI":"10.3233\/faia251127","type":"book-chapter","created":{"date-parts":[[2025,10,22]],"date-time":"2025-10-22T09:52:17Z","timestamp":1761126737000},"source":"Crossref","is-referenced-by-count":0,"title":["Attention Mechanism in Data-Driven Modeling: Empirical Verification"],"prefix":"10.3233","author":[{"ORCID":"https:\/\/orcid.org\/0000-0002-6989-9400","authenticated-orcid":false,"given":"Krzysztof","family":"Patan","sequence":"first","affiliation":[{"name":"University of Zielona G\u00f3ra. Poland"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-2982-6741","authenticated-orcid":false,"given":"Maciej","family":"Patan","sequence":"additional","affiliation":[{"name":"University of Zielona G\u00f3ra. Poland"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-9997-9612","authenticated-orcid":false,"given":"Marek","family":"Kowal","sequence":"additional","affiliation":[{"name":"University of Zielona G\u00f3ra. Poland"}]}],"member":"7437","container-title":["Frontiers in Artificial Intelligence and Applications","ECAI 2025"],"original-title":[],"link":[{"URL":"https:\/\/ebooks.iospress.nl\/pdf\/doi\/10.3233\/FAIA251127","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,10,22]],"date-time":"2025-10-22T09:52:18Z","timestamp":1761126738000},"score":1,"resource":{"primary":{"URL":"https:\/\/ebooks.iospress.nl\/doi\/10.3233\/FAIA251127"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2025,10,21]]},"ISBN":["9781643686318"],"references-count":0,"URL":"https:\/\/doi.org\/10.3233\/faia251127","relation":{},"ISSN":["0922-6389","1879-8314"],"issn-type":[{"value":"0922-6389","type":"print"},{"value":"1879-8314","type":"electronic"}],"subject":[],"published":{"date-parts":[[2025,10,21]]}}}