{"id":{"repo_id":"nps","oai_identifier":"oai:calhoun.nps.edu:10945/73236"},"canonical_url":"https://search.dev.ndltd.org/etd/nps/oai:calhoun.nps.edu:10945/73236","repository":{"repo_id":"nps","name":"Naval Postgraduate School","base_url":"https://calhoun.nps.edu/server/oai/request"},"display":{"title":"REAL-TIME THREAT ASSESSMENT WITH IMPERFECT SENSOR DATA","abstract":"The Department of Defense uses a variety of sensors every day to collect valuable data and information about the adversaries of the United States and its allies. The sensors, however, do not always observe the ground truth because of technological limitations or human errors. This thesis presents a mathematical framework to use sensor-collected data that contain both false negatives and false positives to detect a threat. The first formulation assumes that the sensor operator has intelligence on the threat likelihood at each location, while the second formulation assumes the adversary actively chooses which location to attack to evade sensor detection. In both formulations, we develop a threshold-based policy that points the sensor to the location where it is most likely that an attack is currently taking place to collect more data and raises an alarm if that probability exceeds a location-specific threshold. We use Monte Carlo simulation to evaluate such threshold-based policies based on two conflicting objectives: the probability of detecting a threat in real time and the average time between false alarms. The research findings allow our forces to quantify imperfect sensor data with sound and coherent algorithms rather than relying on ad-hoc assessments and the experiences of subject matter experts.","abstract_html":"The Department of Defense uses a variety of sensors every day to collect valuable data and information about the adversaries of the United States and its allies. The sensors, however, do not always observe the ground truth because of technological limitations or human errors. This thesis presents a mathematical framework to use sensor-collected data that contain both false negatives and false positives to detect a threat. The first formulation assumes that the sensor operator has intelligence on the threat likelihood at each location, while the second formulation assumes the adversary actively chooses which location to attack to evade sensor detection. In both formulations, we develop a threshold-based policy that points the sensor to the location where it is most likely that an attack is currently taking place to collect more data and raises an alarm if that probability exceeds a location-specific threshold. We use Monte Carlo simulation to evaluate such threshold-based policies based on two conflicting objectives: the probability of detecting a threat in real time and the average time between false alarms. The research findings allow our forces to quantify imperfect sensor data with sound and coherent algorithms rather than relying on ad-hoc assessments and the experiences of subject matter experts.","abstract_has_math":false,"creators":["Stanford, Fredrick B."],"institution":"Monterey, CA; Naval Postgraduate School","degree_name":null,"degree_level":null,"degree_discipline":null,"degree_department":"Operations Research (OR)","school":null,"contributors":[],"advisors":["Lin, Kyle Y."],"committee_chairs":[],"committee_members":[],"year":2024,"date_issued":"2024-06","date_published":"2024-06","updated_at":"2026-07-27T20:24:23Z","subjects":[],"languages":[],"rights":["This publication is a work of the U.S. Government as defined in Title 17, United States Code, Section 101. Copyright protection is not available for this work in the United States."],"rights_urls":[],"identifier_entries":[]},"links":{"outbound_url":"https://hdl.handle.net/10945/73236","outbound_label":"Handle","outbound_source":"dc:identifier.uri"},"metadata_groups":[{"id":"people","label":"People","entries":[{"key":"dc:contributor.advisor","label":"Advisor","values":["Lin, Kyle Y."]},{"key":"dc:contributor.department","label":"Department","values":["Operations Research (OR)"]},{"key":"dc:creator","label":"Author","values":["Stanford, Fredrick B."]}]},{"id":"academic_context","label":"Academic Context","entries":[{"key":"dc:date.accessioned","label":"Dc Date Accessioned","values":["2024-08-19T16:38:16Z"]},{"key":"dc:date.available","label":"Dc Date Available","values":["2024-08-19T16:38:16Z"]},{"key":"dc:date.issued","label":"Date","values":["2024-06"]},{"key":"dc:publisher","label":"Institution","values":["Monterey, CA; Naval Postgraduate School"]},{"key":"dc:type","label":"Dc Type","values":["Thesis"]}]},{"id":"language_rights","label":"Language and Rights","entries":[{"key":"dc:rights","label":"Dc Rights","values":["This publication is a work of the U.S. Government as defined in Title 17, United States Code, Section 101. Copyright protection is not available for this work in the United States."]}]},{"id":"identifiers","label":"Identifiers","entries":[{"key":"dc:identifier.uri","label":"Identifier URI","values":["https://hdl.handle.net/10945/73236"]}]},{"id":"additional","label":"Additional Metadata","entries":[{"key":"dc:description.abstract","label":"Abstract","values":["The Department of Defense uses a variety of sensors every day to collect valuable data and information about the adversaries of the United States and its allies. The sensors, however, do not always observe the ground truth because of technological limitations or human errors. This thesis presents a mathematical framework to use sensor-collected data that contain both false negatives and false positives to detect a threat. The first formulation assumes that the sensor operator has intelligence on the threat likelihood at each location, while the second formulation assumes the adversary actively chooses which location to attack to evade sensor detection. In both formulations, we develop a threshold-based policy that points the sensor to the location where it is most likely that an attack is currently taking place to collect more data and raises an alarm if that probability exceeds a location-specific threshold. We use Monte Carlo simulation to evaluate such threshold-based policies based on two conflicting objectives: the probability of detecting a threat in real time and the average time between false alarms. The research findings allow our forces to quantify imperfect sensor data with sound and coherent algorithms rather than relying on ad-hoc assessments and the experiences of subject matter experts."]},{"key":"dc:title","label":"Title","values":["REAL-TIME THREAT ASSESSMENT WITH IMPERFECT SENSOR DATA"]}]}],"canonical_facts":{"dc:contributor.advisor":["Lin, Kyle Y."],"dc:contributor.department":["Operations Research (OR)"],"dc:creator":["Stanford, Fredrick B."],"dc:date.accessioned":["2024-08-19T16:38:16Z"],"dc:date.available":["2024-08-19T16:38:16Z"],"dc:date.issued":["2024-06"],"dc:description.abstract":["The Department of Defense uses a variety of sensors every day to collect valuable data and information about the adversaries of the United States and its allies. The sensors, however, do not always observe the ground truth because of technological limitations or human errors. This thesis presents a mathematical framework to use sensor-collected data that contain both false negatives and false positives to detect a threat. The first formulation assumes that the sensor operator has intelligence on the threat likelihood at each location, while the second formulation assumes the adversary actively chooses which location to attack to evade sensor detection. In both formulations, we develop a threshold-based policy that points the sensor to the location where it is most likely that an attack is currently taking place to collect more data and raises an alarm if that probability exceeds a location-specific threshold. We use Monte Carlo simulation to evaluate such threshold-based policies based on two conflicting objectives: the probability of detecting a threat in real time and the average time between false alarms. The research findings allow our forces to quantify imperfect sensor data with sound and coherent algorithms rather than relying on ad-hoc assessments and the experiences of subject matter experts."],"dc:identifier.uri":["https://hdl.handle.net/10945/73236"],"dc:publisher":["Monterey, CA; Naval Postgraduate School"],"dc:rights":["This publication is a work of the U.S. Government as defined in Title 17, United States Code, Section 101. Copyright protection is not available for this work in the United States."],"dc:title":["REAL-TIME THREAT ASSESSMENT WITH IMPERFECT SENSOR DATA"],"dc:type":["Thesis"]},"updated_at":"2026-07-27T20:24:23Z"}