Non-convex and multi-objective optimization in data mining [Elektronische Ressource] : non-convex and multi-objective optimization for statistical learning and numerical feature engineering / von Ingo Mierswa

Non-Convex and Multi-ObjectiveOptimization in Data MiningNon-Convex and Multi-Objective Optimization forStatistical Learning and Numerical Feature EngineeringDissertationzur Erlangung des Grades einesDoktors der Naturwissenschaftender Technischen Universit¨at Dortmundan der Fakult¨at f¨ur InformatikvonIngo MierswaDortmund20092Tag der mu¨ndlichen Pru¨fung: 27.04.2009Dekan: Prof. Dr. Peter BuchholzGutachter: Prof. Dr. Katharina MorikProf. Dr. Claus Weihs34AcknowledgmentsI would like to express my gratitude to my supervisor, Prof. Dr. Katharina Morik,whose expertise, understanding, patience, and personality added considerably to thisthesis and the last years of my life. I appreciate her vast knowledge and skill and thatshe always gave me the freedom to work on topics of my preference and – maybe evenmore important – to develop my own style of work. During the last years, she becamemore of a mentor and friend than a professor to me.I would like to thank Prof. Dr. Claus Weihs for the assistance he provided and hisvaluablehintshowIcouldimprovemythesis. Itwasreallyhelpfultogetthosecommentson all levels of detail and especially to get these comments from a statistician’s point ofview.I would like to thank the members of the Artificial Intelligence Group, past and present.I find it quite interesting that the time we played soccer almost every day on a parkinglot was also the time everyone of us published most of his work.