{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/the-model-of-an-anomaly-detector-for-hilumi","title":"The model of an anomaly detector for HiLumi LHC magnets based on Recurrent Neural Networks and adaptive quantization","arxiv_id":"1709.09883","date":"2017-09-28","proceeding":null,"authors":["Maciej Wielgosz","Matej Mertik","Andrzej Skoczeń","Ernesto De Matteis"],"abstract":"This paper focuses on an examination of an applicability of Recurrent Neural\nNetwork models for detecting anomalous behavior of the CERN superconducting\nmagnets. In order to conduct the experiments, the authors designed and\nimplemented an adaptive signal quantization algorithm and a custom GRU-based\ndetector and developed a method for the detector parameters selection. Three\ndifferent datasets were used for testing the detector. Two artificially\ngenerated datasets were used to assess the raw performance of the system\nwhereas the 231 MB dataset composed of the signals acquired from HiLumi magnets\nwas intended for real-life experiments and model training. Several different\nsetups of the developed anomaly detection system were evaluated and compared\nwith state-of-the-art OC-SVM reference model operating on the same data. The\nOC-SVM model was equipped with a rich set of feature extractors accounting for\na range of the input signal properties. It was determined in the course of the\nexperiments that the detector, along with its supporting design methodology,\nreaches F1 equal or very close to 1 for almost all test sets. Due to the\nprofile of the data, the best_length setup of the detector turned out to\nperform the best among all five tested configuration schemes of the detection\nsystem. The quantization parameters have the biggest impact on the overall\nperformance of the detector with the best values of input/output grid equal to\n16 and 8, respectively. The proposed solution of the detection significantly\noutperformed OC-SVM-based detector in most of the cases, with much more stable\nperformance across all the datasets.","url_abs":"http://arxiv.org/abs/1709.09883v2","url_pdf":"http://arxiv.org/pdf/1709.09883v2.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"the-model-of-an-anomaly-detector-for-hilumi","repo_url":"https://bitbucket.org/maciekwielgosz/anomaly_detection","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":0,"framework":"none","reach":null}],"tasks":[{"task_slug":"anomaly-detection","task_name":"Anomaly Detection"},{"task_slug":"quantization","task_name":"Quantization"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"syntology_url":null,"atlas_url":null,"mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}