Ensemble Feature Selector for Aviation Accident Analysis and Prevention Sundaram S Meenakshi*, Dr. Kannan S Senthamarai**, Dr. Balamurugan S Appavu Alias*** *Assistant Professor, Pannai College of Engineering & Technology, Tamil Nadu, India **Professor, Department of Computing Science and Engineering, Dhaya College of Engineering, Tamil Nadu, India ***Professor, Department of Information Technoloy, K. L. N. College of Information Technology, Tamil Nadu, India Online published on 2 August, 2016. Abstract Aviation Accidents (AA) are a major public health concern, resulting in a millions of deaths injuries worldwide each year. Aviation Accidents are the leading cause of death and injury. In this work, we applied data mining technologies such feature selection and classification tasks to analyze recorded air accidents during the period 1919–2014, and developed a new data set- Aviation Accident Dataset (AAD) dealing aviation accidents for 1379 accidents with 231 causes as factors for accidents. Among 231 causes the top most cause of accidents are selected by our proposed novel feature selection method “Improved Oscillated Correlation Feature Selection (IOCFS)” and could be used as pre-checklist by the aviation agencies to avoid accidents and improve safety journey. Top Keywords Aviation Accidents (AA), Aviation Accident Dataset (AAD), Aviation Safety Management System, Improved Oscillated Correlation Feature Selection, Classification Algorithm. Top |