[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-84865-en":3,"doc-seo-84865-105":29,"detail-sidebar-cat-0-en-105":82},{"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":21,"language":22,"language_code":23,"site_id":24,"html_lang":23,"table_of_contents":25,"faqs":26,"seo_title":13,"seo_description":14,"update_tm":27,"read_time":28},84865,8796095461564,"Liam","https://ap-avatar.wpscdn.com/davatar_155a257f0dc6eb9ab79c44ca47cae57d",8,"Research & Report","MSCENet A Multi Scale Correlation Enhanced Network for Anomaly Detection","Multivariate time series anomaly detection faces growing difficulty from increasing data complexity and coupled dependencies across temporal scales. MSCENet proposes a spatio-temporal learning framework that jointly captures fine-grained dynamics and cross-series correlations. A fine-grained temporal convolution module uses dilated convolutions to detect both short- and long-term patterns, while a graph-based Mixhop convolution models inter-series relationships across varying scales. A multi-scale gated convolution integrates spatial and temporal attributes to identify subtle multi-scale deviations. Experiments on SMD, PSM, and SWaT demonstrate strong, adaptable performance.","MSCENet: A Multi-Scale Correlation Enhanced Network for Anomaly Detection  \nLong Zhao†, Shixun Ji†, Zhipeng Wang, Member, IEEE, Bin Cheng, Member, IEEE, and Bin He, Senior Member,  \nIEEE  \narXiv :2607 .05864v 1 [ cs .CR] 7 Jul 2026  \nAbstract—In the field of multivariate time series anomaly detection, against the backdrop of increasing data complexity and complex dependencies across multiple temporal scales, traditional methods often struggle to simultaneously capture temporal dynamic features and intricate inter-series correlations. To address this, we propose an innovative framework, MSCENet, which leverages advanced spatio-temporal learning and multi-scale learning techniques to enhance detection accuracy. MSCENet includes a fine-grained temporal convolution module that captures complex temporal dependencies through dilated convolutions, enabling the detection of both short- and long-term patterns. Additionally, the framework models inter-series relationships asa graph structure, using Mixhop graph convolutions to adaptively capture spatial dependencies across varying time scales. To support robust anomaly detection, the multi-scale gated convolution module in MSCENet integrates spatial and temporal attributes through gated mechanisms, facilitating the detection of subtle variations across multiple scales. Experimental evaluations on real-world datasets—SMD, PSM, and SWaT. It provides an adaptable and high-performance solution for anomaly detection in complex time series data environments.  \nIndex Terms—Time series analysis, Multi-scale correlation, Anomaly detection, Machine learning, Deep learning.  \nI. INTRODUCTION  \nTIME series data consists of observations collected at  \nregular intervals, often exhibiting strong temporal dependencies and recurring patterns. It is widely used across sectors such as manufacturing, healthcare, transportation, and network environments, where it includes sensor measurements, network traffic, vehicle motion, and medical information [1],[2] . Time series analysis typically employs regression techniques for forecasting and classification methods for anomaly detection [3], each serving distinct functions. Point anomalies, contextual anomalies, and collective anomalies represent the three primary categories of anomalies in time series data. Point anomalies refer to individual data points that show a marked deviation from the majority of the data. Observations that differ from expected patterns in a given temporal context, often identified through sliding windows, are known as contextual  \nanomalies. On the other hand, collective anomalies occur †These authors contributed equally to this work.  \nThis work was supported in part by the National Natural Science Foundation of China under grants 62573322 and 62088101, by Shanghai Rising-Star Program under grant 24QA2709400, by the Shanghai Chenguang Program under grant 22CGA19, and by the Shanghai Municipal Science and Technology Major Project under grant 2021SHZDZX0100 . (Corresponding author: Bin Cheng.)  \nThe authors are with the Department of Control Science & Engineering, Tongji University, Shanghai 201804, China, and also with the National Key Laboratory of Autonomous Intelligent Unmanned Systems, Shanghai 201203, China (e-mail: [superzhaolong@tongji.edu.cn](superzhaolong@tongji.edu.cn); [jsx@tongji.edu.cn](jsx@tongji.edu.cn);  \n[wangzhipeng@tongji.edu.cn](wangzhipeng@tongji.edu.cn); [bincheng@tongji.edu.cn](bincheng@tongji.edu.cn); [hebin@tongji.edu.cn](hebin@tongji.edu.cn)).  \nwhen clusters of related data points show abnormal behavior in relation to the overall dataset. The variety of anomaly patterns creates significant obstacles for the reliable detection of irregularities within time series data.  \nAs time series data becomes more intricate, particularly with the rise of industrial IoT, dependencies within multivariate time series are magnified [4], [5], complicating the task of anomaly detection. The added reliance on labeled data makes ","cbCaic4gwyzq3pa0","https://ap.wps.com/l/cbCaic4gwyzq3pa0","pdf",5260633,1,14,"English","en",105,"# Introduction\n## Anomaly types in time series\n## Limitations of existing methods\n## Need for multi-scale inter-series correlation modeling","[{\"question\":\"How does MSCENet model relationships between different sequences?\",\"answer\":\"It represents inter-series relationships as a graph and applies Mixhop graph convolutions to adaptively capture spatial (cross-series) dependencies across different time scales.\"}]",1784198905,35,{"code":4,"msg":30,"data":31},"ok",{"site_id":24,"language":23,"slug":32,"title":13,"keywords":33,"description":14,"schema_data":34,"social_meta":77,"head_meta":79,"extra_data":81,"updated_unix":27},"mscenet-a-multi-scale-correlation-enhanced-network-for-anomaly-detection","",{"@graph":35,"@context":76},[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/mscenet-a-multi-scale-correlation-enhanced-network-for-anomaly-detection/84865/",4,{"url":51,"name":13,"@type":54,"author":55,"headline":13,"publisher":57,"fileFormat":60,"inLanguage":23,"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","2026-07-16",true,{"@type":64,"interactionType":65,"userInteractionCount":4},"InteractionCounter",{"@type":66},"ViewAction",{"@type":68,"mainEntity":69},"FAQPage",[70],{"name":71,"@type":72,"acceptedAnswer":73},"How does MSCENet model relationships between different sequences?","Question",{"text":74,"@type":75},"It represents inter-series relationships as a graph and applies Mixhop graph convolutions to adaptively capture spatial (cross-series) dependencies across different time scales.","Answer","https://schema.org",{"og:url":51,"og:type":78,"og:title":13,"og:site_name":58,"og:description":14},"article",{"robots":80,"canonical":51},"index,follow",{"doc_id":7,"site_id":24},{"code":4,"msg":5,"data":83},[84,88,92,96,101,106,111,114,119,122,126],{"id":20,"doc_module":4,"doc_module_name":45,"category_name":85,"show_sort_weight":86,"slug":87},"Story & Novel",90,"story-novel",{"id":46,"doc_module":4,"doc_module_name":45,"category_name":89,"show_sort_weight":90,"slug":91},"Literature",80,"literature",{"id":52,"doc_module":4,"doc_module_name":45,"category_name":93,"show_sort_weight":94,"slug":95},"Exam",70,"exam",{"id":97,"doc_module":4,"doc_module_name":45,"category_name":98,"show_sort_weight":99,"slug":100},5,"Comic",60,"comic",{"id":102,"doc_module":4,"doc_module_name":45,"category_name":103,"show_sort_weight":104,"slug":105},6,"Technology",50,"technology",{"id":107,"doc_module":4,"doc_module_name":45,"category_name":108,"show_sort_weight":109,"slug":110},7,"Healthcare",40,"healthcare",{"id":11,"doc_module":4,"doc_module_name":45,"category_name":12,"show_sort_weight":112,"slug":113},30,"research-report",{"id":115,"doc_module":4,"doc_module_name":45,"category_name":116,"show_sort_weight":117,"slug":118},9,"Religion & Spirituality",20,"religion-spirituality",{"id":117,"doc_module":4,"doc_module_name":45,"category_name":120,"show_sort_weight":117,"slug":121},"World Cup","world-cup",{"id":123,"doc_module":4,"doc_module_name":45,"category_name":124,"show_sort_weight":123,"slug":125},10,"Lifestyle","lifestyle",{"id":127,"doc_module":4,"doc_module_name":45,"category_name":128,"show_sort_weight":97,"slug":129},19,"General","general"]