{"id":{"repo_id":"ets-quebec","oai_identifier":"oai:espace.etsmtl.ca:1704"},"canonical_url":"https://search.ndltd.org/etd/ets-quebec/oai:espace.etsmtl.ca:1704","repository":{"repo_id":"ets-quebec","name":"ETS (Quebec)","base_url":"https://espace.etsmtl.ca/cgi/oai2"},"display":{"title":"Abnormal behavior detection using tensor factorization","abstract":"Real Time Location Systems (RTLS) using RFID is a popular surveillance method for security. However, in an open and dynamic environment, where patterns rarely repeat, it is difficult to implement a model that could analyse all the information generated in real-time and detect abnormal events. In this thesis, we present a new approach to analyze the spatio-temporal information generated by RTL systems. In this approach, discrete events capturing the location of a given person (or object), at a specific time, on a certain day are generated and stored at fixed intervals. Using a latent semantic analysis (LSA) technique based on tensor factorization we extract and leverage latent information contained in real-time streams of multidimensional RFID data represented as these discrete events. One of the main contributions made through this work is a parametric Log-Linear Tensor Factorization (LLTF) model which learns the jointprobability of contextual elements, in which the parameters are the factors of the event tensor. Using an efficient method, based on Nesterov’s accelerated gradient, we obtain a trained model, which is a set of latent factors representing a summarized version of the multidimensional data corresponding to high-level descriptions of each dimension. We evaluated this approach based on LLTF through a series of experiments on synthetic as well as real-life datasets. Upon comparative analysis, results showed that our proposed approach outperformed some state-ofthe-art methods for factor analysis via tensor factorization and abnormal behaviour detection.","abstract_html":"Real Time Location Systems (RTLS) using RFID is a popular surveillance method for security. However, in an open and dynamic environment, where patterns rarely repeat, it is difficult to implement a model that could analyse all the information generated in real-time and detect abnormal events. In this thesis, we present a new approach to analyze the spatio-temporal information generated by RTL systems. In this approach, discrete events capturing the location of a given person (or object), at a specific time, on a certain day are generated and stored at fixed intervals. Using a latent semantic analysis (LSA) technique based on tensor factorization we extract and leverage latent information contained in real-time streams of multidimensional RFID data represented as these discrete events. One of the main contributions made through this work is a parametric Log-Linear Tensor Factorization (LLTF) model which learns the jointprobability of contextual elements, in which the parameters are the factors of the event tensor. Using an efficient method, based on Nesterov’s accelerated gradient, we obtain a trained model, which is a set of latent factors representing a summarized version of the multidimensional data corresponding to high-level descriptions of each dimension. We evaluated this approach based on LLTF through a series of experiments on synthetic as well as real-life datasets. Upon comparative analysis, results showed that our proposed approach outperformed some state-ofthe-art methods for factor analysis via tensor factorization and abnormal behaviour detection.","abstract_has_math":false,"creators":["Shah, Alpa Jayesh"],"institution":"École de technologie supérieure","degree_name":"Mémoire de maîtrise électronique","degree_level":"masters","degree_discipline":null,"degree_department":null,"school":null,"contributors":[],"advisors":[],"committee_chairs":[],"committee_members":[],"year":2016,"date_issued":"2016-05-31","date_published":"2016-05-31","updated_at":"2026-11-18T22:07:07Z","subjects":["Exploration de données (Informatique) Données spatio-temporelles. Apprentissage automatique. Analyse sémantique latente. Factorisation. Calcul tensoriel. Reconnaissance de l'activité humaine (Informatique) Surveillance. anormal, détection, évènement, analyse du comportement, factorisation des tenseurs, apprentissage machine"],"languages":["en"],"rights":[],"rights_urls":[],"identifier_entries":[{"key":"dc:identifier","label":"Identifier","values":["INSERT-YOUR-LIBRARY-ID-HERE1704"],"render_values":[{"text":"INSERT-YOUR-LIBRARY-ID-HERE1704","href":null,"code":true}]}],"person_identifier_entries":[]},"links":{"outbound_url":null,"outbound_label":null,"outbound_source":null,"record_url":null,"record_label":null,"record_kind":null,"record_source":null,"download_url":"https://espace.etsmtl.ca/id/eprint/1704/1/SHAH_Alpa_Jayesh.pdf","download_label":"Download PDF","download_source":"dc:identifier","access_status":null,"embargo_until":null,"access_note":null},"source_record":{"url":"https://espace.etsmtl.ca/cgi/oai2?verb=GetRecord&metadataPrefix=oai_etdms&identifier=oai%3Aespace.etsmtl.ca%3A1704","prefix":"oai_etdms"},"metadata_groups":[{"id":"people","label":"People","entries":[{"key":"dc:creator","label":"Author","values":["Shah, Alpa Jayesh"]}]},{"id":"academic_context","label":"Academic Context","entries":[{"key":"dc:date","label":"Dc Date","values":["2016-05-31"]},{"key":"dc:publisher","label":"Institution","values":["École de technologie supérieure"]},{"key":"dc:type","label":"Dc Type","values":["Electronic Thesis or Dissertation"]},{"key":"thesis:degree_level","label":"Degree Level","values":["masters"]},{"key":"thesis:degree_name","label":"Degree Name","values":["Mémoire de maîtrise électronique"]},{"key":"thesis:institution_name","label":"Thesis Institution Name","values":["École de technologie supérieure"]}]},{"id":"subjects_keywords","label":"Subjects and Keywords","entries":[{"key":"dc:subject","label":"Dc Subject","values":["Exploration de données (Informatique) Données spatio-temporelles. Apprentissage automatique. Analyse sémantique latente. Factorisation. Calcul tensoriel. Reconnaissance de l'activité humaine (Informatique) Surveillance. anormal, détection, évènement, analyse du comportement, factorisation des tenseurs, apprentissage machine"]}]},{"id":"language_rights","label":"Language and Rights","entries":[{"key":"dc:language","label":"Dc Language","values":["en"]}]},{"id":"identifiers","label":"Identifiers","entries":[{"key":"dc:identifier","label":"Identifier","values":["INSERT-YOUR-LIBRARY-ID-HERE1704","https://espace.etsmtl.ca/id/eprint/1704/1/SHAH_Alpa_Jayesh.pdf","https://espace.etsmtl.ca/id/eprint/1704/2/SHAH_Alpa_Jayesh-web.pdf"]}]},{"id":"additional","label":"Additional Metadata","entries":[{"key":"dc:description","label":"Description","values":["Real Time Location Systems (RTLS) using RFID is a popular surveillance method for security. However, in an open and dynamic environment, where patterns rarely repeat, it is difficult to implement a model that could analyse all the information generated in real-time and detect abnormal events. In this thesis, we present a new approach to analyze the spatio-temporal information generated by RTL systems. In this approach, discrete events capturing the location of a given person (or object), at a specific time, on a certain day are generated and stored at fixed intervals. Using a latent semantic analysis (LSA) technique based on tensor factorization we extract and leverage latent information contained in real-time streams of multidimensional RFID data represented as these discrete events. One of the main contributions made through this work is a parametric Log-Linear Tensor Factorization (LLTF) model which learns the jointprobability of contextual elements, in which the parameters are the factors of the event tensor. Using an efficient method, based on Nesterov’s accelerated gradient, we obtain a trained model, which is a set of latent factors representing a summarized version of the multidimensional data corresponding to high-level descriptions of each dimension. We evaluated this approach based on LLTF through a series of experiments on synthetic as well as real-life datasets. Upon comparative analysis, results showed that our proposed approach outperformed some state-ofthe-art methods for factor analysis via tensor factorization and abnormal behaviour detection."]},{"key":"dc:format","label":"Dc Format","values":["application/pdf"]},{"key":"dc:title","label":"Title","values":["Abnormal behavior detection using tensor factorization"]}]}],"canonical_facts":{"dc:creator":["Shah, Alpa Jayesh"],"dc:date":["2016-05-31"],"dc:description":["Real Time Location Systems (RTLS) using RFID is a popular surveillance method for security. However, in an open and dynamic environment, where patterns rarely repeat, it is difficult to implement a model that could analyse all the information generated in real-time and detect abnormal events. In this thesis, we present a new approach to analyze the spatio-temporal information generated by RTL systems. In this approach, discrete events capturing the location of a given person (or object), at a specific time, on a certain day are generated and stored at fixed intervals. Using a latent semantic analysis (LSA) technique based on tensor factorization we extract and leverage latent information contained in real-time streams of multidimensional RFID data represented as these discrete events. One of the main contributions made through this work is a parametric Log-Linear Tensor Factorization (LLTF) model which learns the jointprobability of contextual elements, in which the parameters are the factors of the event tensor. Using an efficient method, based on Nesterov’s accelerated gradient, we obtain a trained model, which is a set of latent factors representing a summarized version of the multidimensional data corresponding to high-level descriptions of each dimension. We evaluated this approach based on LLTF through a series of experiments on synthetic as well as real-life datasets. Upon comparative analysis, results showed that our proposed approach outperformed some state-ofthe-art methods for factor analysis via tensor factorization and abnormal behaviour detection."],"dc:format":["application/pdf"],"dc:identifier":["INSERT-YOUR-LIBRARY-ID-HERE1704","https://espace.etsmtl.ca/id/eprint/1704/1/SHAH_Alpa_Jayesh.pdf","https://espace.etsmtl.ca/id/eprint/1704/2/SHAH_Alpa_Jayesh-web.pdf"],"dc:language":["en"],"dc:publisher":["École de technologie supérieure"],"dc:subject":["Exploration de données (Informatique) Données spatio-temporelles. Apprentissage automatique. Analyse sémantique latente. Factorisation. Calcul tensoriel. Reconnaissance de l'activité humaine (Informatique) Surveillance. anormal, détection, évènement, analyse du comportement, factorisation des tenseurs, apprentissage machine"],"dc:title":["Abnormal behavior detection using tensor factorization"],"dc:type":["Electronic Thesis or Dissertation"],"thesis:degree_level":["masters"],"thesis:degree_name":["Mémoire de maîtrise électronique"],"thesis:institution_name":["École de technologie supérieure"]},"updated_at":"2026-11-18T22:07:07Z"}