{"id":{"repo_id":"marshall","oai_identifier":"oai:mds.marshall.edu:etd-2818"},"canonical_url":"https://search.ndltd.org/etd/marshall/oai:mds.marshall.edu:etd-2818","repository":{"repo_id":"marshall","name":"Marshall University","base_url":"https://mds.marshall.edu/do/oai/"},"display":{"title":"Identifying hazardous patterns in MSHA data using random forests","abstract":"Mining safety and health in the US can be better understood through the application of machine learning techniques to data collected by the Mine Safety and Health Administration (MSHA). By identifying hazardous conditions that could lead to accidents before they occur, valuable insights can be gained by MSHA, mining operators, and miners. In this study, we propose using a Random Forest machine learning model to predict whether a given mining violation will lead to an accident, and if so, whether it will be fatal or non-fatal. To achieve this, the model is trained on MSHA violation data and the sum of scheduled accident charges within 35 days of the violation. We experiment with different predictive models using varying data columns, training set sizes, prediction classes, and hyperparameters to achieve a reliable prediction. One of the challenges in generating these models is accurately predicting the sparse class of accidents, as opposed to the abundant class of no accidents. To address this, we propose utilizing sample minimizing to balance the false negative and false positive rate and create a more accurate predictive model. Our results demonstrate, with a high degree of confidence, the potential for machine learning to improve mine safety and health by identifying hazardous conditions and mitigating the risk of accidents.","abstract_html":"Mining safety and health in the US can be better understood through the application of machine learning techniques to data collected by the Mine Safety and Health Administration (MSHA). By identifying hazardous conditions that could lead to accidents before they occur, valuable insights can be gained by MSHA, mining operators, and miners. In this study, we propose using a Random Forest machine learning model to predict whether a given mining violation will lead to an accident, and if so, whether it will be fatal or non-fatal. To achieve this, the model is trained on MSHA violation data and the sum of scheduled accident charges within 35 days of the violation. We experiment with different predictive models using varying data columns, training set sizes, prediction classes, and hyperparameters to achieve a reliable prediction. One of the challenges in generating these models is accurately predicting the sparse class of accidents, as opposed to the abundant class of no accidents. To address this, we propose utilizing sample minimizing to balance the false negative and false positive rate and create a more accurate predictive model. Our results demonstrate, with a high degree of confidence, the potential for machine learning to improve mine safety and health by identifying hazardous conditions and mitigating the risk of accidents.","abstract_has_math":false,"creators":["Milam, Olivia"],"institution":null,"degree_name":"Computer Science","degree_level":"Thesis","degree_discipline":"College of Engineering and Computer Sciences","degree_department":null,"school":null,"contributors":["Dr. James D. McIntosh, Committee Chairperson","Dr. Tanvir Irfan Chowdhury","Dr. Ammar Alzarrad"],"advisors":[],"committee_chairs":[],"committee_members":[],"year":null,"date_issued":"","date_published":null,"updated_at":"2026-11-21T19:58:01Z","subjects":["Computer Engineering","Engineering","Risk Analysis","Social and Behavioral Sciences"],"languages":[],"rights":[],"rights_urls":[],"identifier_entries":[],"person_identifier_entries":[]},"links":{"outbound_url":"https://mds.marshall.edu/etd/1813","outbound_label":"Repository record","outbound_source":"dc:identifier","record_url":"https://mds.marshall.edu/etd/1813","record_label":"Repository record","record_kind":"record","record_source":"dc:identifier","download_url":null,"download_label":null,"download_source":null,"access_status":null,"embargo_until":null,"access_note":null},"source_record":{"url":"https://mds.marshall.edu/do/oai/?verb=GetRecord&metadataPrefix=oai_etdms&identifier=oai%3Amds.marshall.edu%3Aetd-2818","prefix":"oai_etdms"},"metadata_groups":[{"id":"people","label":"People","entries":[{"key":"dc:contributor","label":"Contributor","values":["Dr. James D. McIntosh, Committee Chairperson","Dr. Tanvir Irfan Chowdhury","Dr. Ammar Alzarrad"]},{"key":"dc:creator","label":"Author","values":["Milam, Olivia"]}]},{"id":"academic_context","label":"Academic Context","entries":[{"key":"dc:date.available","label":"Dc Date Available","values":["2023-07-31T07:00:00Z"]},{"key":"dc:date.created","label":"Dc Date Created","values":["2023-01-01T08:00:00Z"]},{"key":"thesis:degree_discipline","label":"Discipline","values":["College of Engineering and Computer Sciences"]},{"key":"thesis:degree_level","label":"Degree Level","values":["Thesis"]},{"key":"thesis:degree_name","label":"Degree Name","values":["Computer Science"]}]},{"id":"subjects_keywords","label":"Subjects and Keywords","entries":[{"key":"dc:subject","label":"Dc Subject","values":["Computer Engineering","Engineering","Risk Analysis","Social and Behavioral Sciences"]}]},{"id":"identifiers","label":"Identifiers","entries":[{"key":"dc:identifier","label":"Identifier","values":["https://mds.marshall.edu/etd/1813"]}]},{"id":"additional","label":"Additional Metadata","entries":[{"key":"dc:description.abstract","label":"Abstract","values":["Mining safety and health in the US can be better understood through the application of machine learning techniques to data collected by the Mine Safety and Health Administration (MSHA). By identifying hazardous conditions that could lead to accidents before they occur, valuable insights can be gained by MSHA, mining operators, and miners. In this study, we propose using a Random Forest machine learning model to predict whether a given mining violation will lead to an accident, and if so, whether it will be fatal or non-fatal. To achieve this, the model is trained on MSHA violation data and the sum of scheduled accident charges within 35 days of the violation. We experiment with different predictive models using varying data columns, training set sizes, prediction classes, and hyperparameters to achieve a reliable prediction. One of the challenges in generating these models is accurately predicting the sparse class of accidents, as opposed to the abundant class of no accidents. To address this, we propose utilizing sample minimizing to balance the false negative and false positive rate and create a more accurate predictive model. Our results demonstrate, with a high degree of confidence, the potential for machine learning to improve mine safety and health by identifying hazardous conditions and mitigating the risk of accidents."]},{"key":"dc:title","label":"Title","values":["Identifying hazardous patterns in MSHA data using random forests"]}]}],"canonical_facts":{"dc:contributor":["Dr. James D. McIntosh, Committee Chairperson","Dr. Tanvir Irfan Chowdhury","Dr. Ammar Alzarrad"],"dc:creator":["Milam, Olivia"],"dc:date.available":["2023-07-31T07:00:00Z"],"dc:date.created":["2023-01-01T08:00:00Z"],"dc:description.abstract":["Mining safety and health in the US can be better understood through the application of machine learning techniques to data collected by the Mine Safety and Health Administration (MSHA). By identifying hazardous conditions that could lead to accidents before they occur, valuable insights can be gained by MSHA, mining operators, and miners. In this study, we propose using a Random Forest machine learning model to predict whether a given mining violation will lead to an accident, and if so, whether it will be fatal or non-fatal. To achieve this, the model is trained on MSHA violation data and the sum of scheduled accident charges within 35 days of the violation. We experiment with different predictive models using varying data columns, training set sizes, prediction classes, and hyperparameters to achieve a reliable prediction. One of the challenges in generating these models is accurately predicting the sparse class of accidents, as opposed to the abundant class of no accidents. To address this, we propose utilizing sample minimizing to balance the false negative and false positive rate and create a more accurate predictive model. Our results demonstrate, with a high degree of confidence, the potential for machine learning to improve mine safety and health by identifying hazardous conditions and mitigating the risk of accidents."],"dc:identifier":["https://mds.marshall.edu/etd/1813"],"dc:subject":["Computer Engineering","Engineering","Risk Analysis","Social and Behavioral Sciences"],"dc:title":["Identifying hazardous patterns in MSHA data using random forests"],"thesis:degree_discipline":["College of Engineering and Computer Sciences"],"thesis:degree_level":["Thesis"],"thesis:degree_name":["Computer Science"]},"updated_at":"2026-11-21T19:58:01Z"}