[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-117890-en":3,"doc-seo-117890-105":30,"detail-sidebar-cat-0-en-105":92},{"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":20,"is_deleted":4,"is_public":20,"is_downloadable":20,"audit_status":20,"page_count":21,"language":22,"language_code":23,"site_id":24,"html_lang":23,"table_of_contents":25,"faqs":26,"seo_title":27,"seo_description":14,"update_tm":28,"read_time":29},117890,8796095360427,"Lucas Martin","https://ap-avatar.wpscdn.com/davatar_994ba38a5ba835b3df7d355c54d3ed8d",8,"Research & Report","Autonomous Anomaly Detection Via Physics-Regularized Machine Learning","Manual inspection of telemetry data for anomaly detection is time-consuming and hard to scale when multi-signal systems stream extensive information. This report frames autonomous anomaly detection as regression and classification: reconstruct signals, then classify reconstruction error as anomalous or nominal. It evaluates physics-informed regression by incorporating target physics, improving reconstruction error over purely data-driven modeling. For classification, it studies cluster-based thresholding with pruning, achieving stronger performance than non-parametric dynamic thresholds.","Copyright by  \nFelipe Giraldo-Grueso 2023  \nAutonomous Anomaly Detection Via Physics-Regularized Machine Learning  \nSUPERVISING COMMITTEE:  \nRenato Zanetti, Supervisor  \nMaruthi Akella  \nAutonomous Anomaly Detection Via Physics-Regularized Machine Learning  \nby  \nFelipe Giraldo-Grueso, B.S .  \nREPORT  \nPresented to the Faculty of the Graduate School of The University of Texas at Austin in Partial Fulfillment of the Requirements for the Degree of  \nMASTER OF SCIENCE IN ENGINEERING  \nTHE UNIVERSITY OF TEXAS AT AUSTIN  \nMay 2023  \nDedicated to Pascal  \nAcknowledgments  \nI would like to express my gratitude to Professor Renato Zanetti, my research supervisor, for his guidance, encouragement, and useful critiques of this research work.  \nI would also like to extend my appreciation to Professor Maruthi Akella for participating as a report reviewer, and to the Department of Aerospace Engineering and Engineering Mechanics at The University of Texas at Austin for their help in offering me the resources needed to complete this work.  \nFinally, I deeply and sincerely thank my family and friends for their continuous love, help, and support.  \nThis material is based on research sponsored by Air Force Research Laboratory (AFRL) under agreement number FA9453-21-1-0045 .  \nAutonomous Anomaly Detection Via Physics-Regularized Machine Learning  \nFelipe Giraldo-Grueso, M.S.E.  \nThe University of Texas at Austin, 2023  \nSupervisor: Renato Zanetti  \nManual inspection of telemetry data in the search for anomalies is a time-consuming threat detection technique. Most multi-signal systems send back extensive data that a single person cannot easily monitor in real time. Machine learning techniques that autonomously scan data and flag anomalies are attractive alternatives. The autonomous anomaly detection problem can be divided into two sub-problems: regression analysis and a classification process. In the regression analysis, a machine learning model is trained to reconstruct a given signal, and the classification process categorizes thereconstruction error as anomalous or nominal. This report examines the autonomous anomaly detection problem and proposes improvements to both the regression and classification sub-problems. With regard to the regression analysis, it was found that including the physics of the target signal in the machine learning model yielded a lower reconstruction error compared to a purely datadriven model. The classification approaches studied showed that cluster-based  \nthresholding techniques accompanied by a pruning procedure can outperform non-parametric dynamic thresholds.  \nTable of Contents  \nAcknowledgments v  \nAbstract vi  \nList of Tables x  \nList of Figures xi  \nChapter 1 . Introduction 1  \n1.1 Related Work ........................... 3  \nChapter 2 . Data Preparation 5  \n2.1 Regression ............................. 5  \n2.2 Classification ............................ 7  \n2.2.1 Real Anomalies ....................... 7  \n2.2.2 Synthetic Anomalies .................... 8  \n2.3 Summary of Data ......................... 10  \nChapter 3 . Learning Framework 12  \n3.1 ACE dataset ............................ 12  \n3.1.1 Autoencoders ........................ 12  \n3.1.2 Loss functions ....................... 13  \n3.1.3 Bias-Variance Trade-off .................. 15  \n3.1.4 Curriculum Regularization ................ 16  \n3.2 SMAP/MSL dataset ....................... 17  \n3.2.1 Long Short-Term Memory Networks ........... 17  \n3.2.2 Gaussian Assumption ................... 19  \n3.2.3 K-means clustering ..................... 19  \n3.2.4 Pruning procedure ..................... 20  \n3.3 ACEC dataset ........................... 21  \nChapter 4 . Training and Testing 22  \n4.1 Metrics ............................... 22  \nChapter 5 . Results 25  \n5.1 Regression ............................. 25  \n5.1.1 Physics Regularization ................... 25  \n5.1.2 Curriculum Regularization ................ 28  \n5.2 Classification ............................ 30","cbCaipS6HYc6yYNS","https://ap.wps.com/l/cbCaipS6HYc6yYNS","pdf",6931851,1,65,"English","en",105,"# Acknowledgments\n# Abstract\n# List of Tables\n# List of Figures\n# Chapter 1. Introduction\n## 1.1 Related Work\n# Chapter 2. Data Preparation\n## 2.1 Regression\n## 2.2 Classification\n## 2.2.1 Real Anomalies\n## 2.2.2 Synthetic Anomalies\n## 2.3 Summary of Data\n# Chapter 3. Learning Framework\n## 3.1 ACE dataset\n## 3.2 SMAP/MSL dataset\n## 3.3 ACEC dataset\n# Chapter 4. Training and Testing\n## 4.1 Metrics\n# Chapter 5. Results\n## 5.1 Regression\n## 5.2 Classification\n# Chapter 6. Conclusion and Future Work\n# Appendices\n## Appendix A. Reconstructions\n# Bibliography","[{\"question\":\"How is the autonomous anomaly detection task structured in the report?\",\"answer\":\"The task is split into regression and classification. A model reconstructs the signal, and the classification step labels reconstruction error as anomalous or nominal.\"},{\"question\":\"What is the impact of physics regularization on reconstruction error?\",\"answer\":\"Including the physics of the target signal in the machine learning model produces a lower reconstruction error than a purely data-driven approach.\"},{\"question\":\"Which classification strategy is shown to perform well?\",\"answer\":\"Cluster-based thresholding combined with a pruning procedure can outperform non-parametric dynamic thresholds, according to the studied approaches.\"}]","Autonomous Anomaly Detection Via Physics-Regularized Machine Learning | PDF",1785680201,164,{"code":4,"msg":31,"data":32},"ok",{"site_id":24,"language":23,"slug":33,"title":13,"keywords":34,"description":14,"schema_data":35,"social_meta":87,"head_meta":89,"extra_data":91,"updated_unix":28},"autonomous-anomaly-detection-via-physics-regularized-machine-learning","",{"@graph":36,"@context":86},[37,54,69],{"@type":38,"itemListElement":39},"BreadcrumbList",[40,44,48,51],{"item":41,"name":42,"@type":43,"position":20},"https://docshare.wps.com","Home","ListItem",{"item":45,"name":46,"@type":43,"position":47},"https://docshare.wps.com/document/","Document",2,{"item":49,"name":12,"@type":43,"position":50},"https://docshare.wps.com/document/research-report/",3,{"item":52,"name":13,"@type":43,"position":53},"https://docshare.wps.com/document/autonomous-anomaly-detection-via-physics-regularized-machine-learning/117890/",4,{"url":52,"name":13,"@type":55,"author":56,"headline":13,"publisher":58,"fileFormat":61,"inLanguage":23,"description":14,"dateModified":62,"datePublished":63,"encodingFormat":61,"isAccessibleForFree":64,"interactionStatistic":65},"DigitalDocument",{"name":9,"@type":57},"Person",{"url":41,"name":59,"@type":60},"DocShare","Organization","application/pdf","2026-08-05","2026-08-02",true,{"@type":66,"interactionType":67,"userInteractionCount":20},"InteractionCounter",{"@type":68},"ViewAction",{"@type":70,"mainEntity":71},"FAQPage",[72,78,82],{"name":73,"@type":74,"acceptedAnswer":75},"How is the autonomous anomaly detection task structured in the report?","Question",{"text":76,"@type":77},"The task is split into regression and classification. A model reconstructs the signal, and the classification step labels reconstruction error as anomalous or nominal.","Answer",{"name":79,"@type":74,"acceptedAnswer":80},"What is the impact of physics regularization on reconstruction error?",{"text":81,"@type":77},"Including the physics of the target signal in the machine learning model produces a lower reconstruction error than a purely data-driven approach.",{"name":83,"@type":74,"acceptedAnswer":84},"Which classification strategy is shown to perform well?",{"text":85,"@type":77},"Cluster-based thresholding combined with a pruning procedure can outperform non-parametric dynamic thresholds, according to the studied approaches.","https://schema.org",{"og:url":52,"og:type":88,"og:title":13,"og:site_name":59,"og:description":14},"article",{"robots":90,"canonical":52},"index,follow",{"doc_id":7,"site_id":24},{"code":4,"msg":5,"data":93},[94,98,102,106,111,116,121,124,129,132,136],{"id":20,"doc_module":4,"doc_module_name":46,"category_name":95,"show_sort_weight":96,"slug":97},"Story & Novel",90,"story-novel",{"id":47,"doc_module":4,"doc_module_name":46,"category_name":99,"show_sort_weight":100,"slug":101},"Literature",80,"literature",{"id":53,"doc_module":4,"doc_module_name":46,"category_name":103,"show_sort_weight":104,"slug":105},"Exam",70,"exam",{"id":107,"doc_module":4,"doc_module_name":46,"category_name":108,"show_sort_weight":109,"slug":110},5,"Comic",60,"comic",{"id":112,"doc_module":4,"doc_module_name":46,"category_name":113,"show_sort_weight":114,"slug":115},6,"Technology",50,"technology",{"id":117,"doc_module":4,"doc_module_name":46,"category_name":118,"show_sort_weight":119,"slug":120},7,"Healthcare",40,"healthcare",{"id":11,"doc_module":4,"doc_module_name":46,"category_name":12,"show_sort_weight":122,"slug":123},30,"research-report",{"id":125,"doc_module":4,"doc_module_name":46,"category_name":126,"show_sort_weight":127,"slug":128},9,"Religion & Spirituality",20,"religion-spirituality",{"id":127,"doc_module":4,"doc_module_name":46,"category_name":130,"show_sort_weight":127,"slug":131},"World Cup","world-cup",{"id":133,"doc_module":4,"doc_module_name":46,"category_name":134,"show_sort_weight":133,"slug":135},10,"Lifestyle","lifestyle",{"id":137,"doc_module":4,"doc_module_name":46,"category_name":138,"show_sort_weight":107,"slug":139},19,"General","general"]