[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-121444-en":3,"doc-seo-121444-105":29,"detail-sidebar-cat-0-en-105":90},{"code":4,"msg":5,"data":6},0,"success",{"doc_id":7,"user_id":8,"nickname":9,"user_avatar":10,"doc_module":4,"category_id":11,"category_name":12,"doc_title":13,"doc_description":14,"doc_content":15,"file_id":16,"file_url":17,"file_type":18,"file_size":19,"view_count":4,"is_deleted":4,"is_public":20,"is_downloadable":20,"audit_status":20,"page_count":11,"language":21,"language_code":22,"site_id":23,"html_lang":22,"table_of_contents":24,"faqs":25,"seo_title":26,"seo_description":14,"update_tm":27,"read_time":28},121444,13056703019662,"Evangeline","https://ap-avatar.wpscdn.com/avatar/be000253a8e92610077?_k=1778726343310543188",8,"Research & Report","Exploring Interpretable Machine Learning for Modeling Nonlinear Dynamic Systems - Conference Paper","This paper analyzes interpretable machine learning (IML) methods for modeling nonlinear dynamic systems, with emphasis on restoring-force representation. The approach uses neural networks that take displacement and velocity as inputs, and employs a parallel network configured with the proposed architecture and random initialization to mirror the designed model. Multiple training and testing cycles on identical datasets support robust statistical validation. Results show that interpretable artificial neural networks achieve higher accuracy and more consistent outcomes than randomly initialized counterparts, supporting IML as a practical path for improved physical modeling and generalization in engineering applications.","University for Business and Technology in Kosovo  \nUBT Knowledge Center  \n\n| UBT International Conference | 2024 UBT International Conference |\n| --- | --- |\n| Oct 25th, 9:00 AM-Oct 27th, 6:00 PM\u003Cbr>Exploring Interpretable Machine Learning for Modeling Nonlinear Dynamic Systems\u003Cbr>Liron Morina\u003Cbr>University of Prishtina, [liron.morina@uni-pr.edu](liron.morina@uni-pr.edu)\u003Cbr>Rina Peja\u003Cbr>University for Business and Technology-UBT, [rinapeja1@gmail.com](rinapeja1@gmail.com)\u003Cbr>Getoar Vranovci\u003Cbr>University for Business and Technology-UBT, [getoarvranovci@gmail.com](getoarvranovci@gmail.com)\u003Cbr>Follow this and additional works at: [https://knowledgecenter.ubt-uni.net/conference](https://knowledgecenter.ubt-uni.net/conference)\u003Cbr> Part of the Engineering Commons |  |\n\nRecommended Citation  \nMorina, Liron; Peja, Rina; and Vranovci, Getoar, \"Exploring Interpretable Machine Learning for Modeling Nonlinear Dynamic Systems\" (2024) . UBT International Conference. 15.  \n[https://knowledgecenter.ubt-uni.net/conference/2024UBTIC/CEIE/15](https://knowledgecenter.ubt-uni.net/conference/2024UBTIC/CEIE/15)  \nThis Event is brought to you for free and open access by the Publication and Journals at UBT Knowledge Center. It has been accepted for inclusion in UBT International Conference by an authorized administrator of UBT Knowledge Center. For more information, please contact [knowledge.center@ubt-uni.net](knowledge.center@ubt-uni.net).  \nExploring Interpretable Machine Learning for Modeling Nonlinear Dynamic Systems  \nLiron Morina 1, Rina Peja2, Getoar Vranovci3  \n1 University of Prishtina, Rr.“George Bush”, Nr. 31, 10000 Prishtine, Rep. Kosoves  \n2 University of Business and Technology, Lagjia Kalabria, 10000 Prishtine, Rep. Kosoves  \n3 University of Prishtina, Rr.“George Bush”, Nr. 31, 10000 Prishtine, Rep. Kosoves  \nAbstract. This paper provides an analysis of the interpretable machine learning (IML) approach applied to various nonlinear dynamic systems. The study focuses on modeling the restoring force by neural networks with two input values: displacement and velocity. A parallel neural network with the proposed architecture, initialized randomly, mirrors the designed model. Both models undergo multiple training and testing cycles on identical datasets to ensure robust statistical validation. The findings demonstrate that interpretable artificial neural networks outperform randomly initialized models in terms of accuracy and result consistency.  \nThis research provides valuable insights into the application of IML techniques for advancing the modeling capabilities of nonlinear dynamic systems.  \nKeywords: Nonlinear dynamic systems, Interpretable Machine Learning, Artificial Neural Networks, Initialization  \n1 Introduction  \nArtificial neural networks (ANNs) have gained significant popularity in the twenty-first century. As described in [1], ANNs consist of \"an interconnected assembly of simple processing elements, units, or nodes, whose functionality is loosely based on the animal neuron.\" Their effectiveness stems from the connections between these nodes, known as weights, which are established through a learning process and indicate the strength of relationships among the nodes. Thanks to their efficiency in pattern recognition and their ability to predict outcomes for similar inputs, ANNs are widely utilized across various scientific disciplines. However, researchers exercise caution when employing ANNs due to certain limitations. One key concern is that the weights and biases often lack clear physical interpretation. Additionally, the non-uniqueness of the trained parameter values can complicate matters, as researchers must determine which set of parameters is most suitable for their specific application.  \nDespite the opacity of neural networks, their proven effectiveness leads to the argument that researchers may need to accept these less transparent aspects, especially when compared to traditional computational tools [2] . Addressin","cbCairGKO1d4Ii7H","https://ap.wps.com/l/cbCairGKO1d4Ii7H","pdf",458081,1,"English","en",105,"# Abstract\n# 1 Introduction\n## Artificial neural networks and interpretability challenges\n## Interpretable ANNs and physically meaningful initialization\n# 2 Literature review\n## Universal approximation theorem and interpretability foundations","[{\"question\":\"What nonlinear dynamic systems problem does the paper target?\",\"answer\":\"It targets modeling the restoring force in nonlinear dynamic systems by learning relationships from displacement and velocity inputs using neural networks.\"},{\"question\":\"How does the study compare interpretable models with baseline models?\",\"answer\":\"It trains and tests an interpretable artificial neural network and a parallel network using the proposed architecture with random initialization, running multiple training/testing cycles on identical datasets to compare performance and variance.\"},{\"question\":\"How are the training and testing datasets constructed?\",\"answer\":\"Training uses data generated from sine excitation, while evaluation uses a test dataset derived from swept-sine excitation to assess generalization capability.\"}]","Exploring Interpretable Machine Learning for Modeling Nonlinear Dynamic Systems - Conference Paper | PDF",1785735692,20,{"code":4,"msg":30,"data":31},"ok",{"site_id":23,"language":22,"slug":32,"title":13,"keywords":33,"description":14,"schema_data":34,"social_meta":85,"head_meta":87,"extra_data":89,"updated_unix":27},"exploring-interpretable-machine-learning-for-modeling-nonlinear-dynamic-systems-conference-paper","",{"@graph":35,"@context":84},[36,53,67],{"@type":37,"itemListElement":38},"BreadcrumbList",[39,43,47,50],{"item":40,"name":41,"@type":42,"position":20},"https://docshare.wps.com","Home","ListItem",{"item":44,"name":45,"@type":42,"position":46},"https://docshare.wps.com/document/","Document",2,{"item":48,"name":12,"@type":42,"position":49},"https://docshare.wps.com/document/research-report/",3,{"item":51,"name":13,"@type":42,"position":52},"https://docshare.wps.com/document/exploring-interpretable-machine-learning-for-modeling-nonlinear-dynamic-systems-conference-paper/121444/",4,{"url":51,"name":13,"@type":54,"author":55,"headline":13,"publisher":57,"fileFormat":60,"inLanguage":22,"description":14,"dateModified":61,"datePublished":61,"encodingFormat":60,"isAccessibleForFree":62,"interactionStatistic":63},"DigitalDocument",{"name":9,"@type":56},"Person",{"url":40,"name":58,"@type":59},"DocShare","Organization","application/pdf","2026-08-03",true,{"@type":64,"interactionType":65,"userInteractionCount":4},"InteractionCounter",{"@type":66},"ViewAction",{"@type":68,"mainEntity":69},"FAQPage",[70,76,80],{"name":71,"@type":72,"acceptedAnswer":73},"What nonlinear dynamic systems problem does the paper target?","Question",{"text":74,"@type":75},"It targets modeling the restoring force in nonlinear dynamic systems by learning relationships from displacement and velocity inputs using neural networks.","Answer",{"name":77,"@type":72,"acceptedAnswer":78},"How does the study compare interpretable models with baseline models?",{"text":79,"@type":75},"It trains and tests an interpretable artificial neural network and a parallel network using the proposed architecture with random initialization, running multiple training/testing cycles on identical datasets to compare performance and variance.",{"name":81,"@type":72,"acceptedAnswer":82},"How are the training and testing datasets constructed?",{"text":83,"@type":75},"Training uses data generated from sine excitation, while evaluation uses a test dataset derived from swept-sine excitation to assess generalization capability.","https://schema.org",{"og:url":51,"og:type":86,"og:title":13,"og:site_name":58,"og:description":14},"article",{"robots":88,"canonical":51},"index,follow",{"doc_id":7,"site_id":23},{"code":4,"msg":5,"data":91},[92,96,100,104,109,114,119,122,126,129,133],{"id":20,"doc_module":4,"doc_module_name":45,"category_name":93,"show_sort_weight":94,"slug":95},"Story & Novel",90,"story-novel",{"id":46,"doc_module":4,"doc_module_name":45,"category_name":97,"show_sort_weight":98,"slug":99},"Literature",80,"literature",{"id":52,"doc_module":4,"doc_module_name":45,"category_name":101,"show_sort_weight":102,"slug":103},"Exam",70,"exam",{"id":105,"doc_module":4,"doc_module_name":45,"category_name":106,"show_sort_weight":107,"slug":108},5,"Comic",60,"comic",{"id":110,"doc_module":4,"doc_module_name":45,"category_name":111,"show_sort_weight":112,"slug":113},6,"Technology",50,"technology",{"id":115,"doc_module":4,"doc_module_name":45,"category_name":116,"show_sort_weight":117,"slug":118},7,"Healthcare",40,"healthcare",{"id":11,"doc_module":4,"doc_module_name":45,"category_name":12,"show_sort_weight":120,"slug":121},30,"research-report",{"id":123,"doc_module":4,"doc_module_name":45,"category_name":124,"show_sort_weight":28,"slug":125},9,"Religion & Spirituality","religion-spirituality",{"id":28,"doc_module":4,"doc_module_name":45,"category_name":127,"show_sort_weight":28,"slug":128},"World Cup","world-cup",{"id":130,"doc_module":4,"doc_module_name":45,"category_name":131,"show_sort_weight":130,"slug":132},10,"Lifestyle","lifestyle",{"id":134,"doc_module":4,"doc_module_name":45,"category_name":135,"show_sort_weight":105,"slug":136},19,"General","general"]