讲座人:Dr. Vikrant Vaze达特茅斯学院
讲座时间:2016年6月7日14:00-15:30
讲座地点:address
讲座内容:Abstract: This talk will focus on data-driven and statistical modeling techniques using large real-world datasets. We will focus on three different case studies in the context of the US air transportation system. The US National Aviation System (NAS) is a complex system with multiple, interacting agents including airlines, passengers, and regulators, each with somewhat different objectives and incentives. These interactions determine the state of the system in terms of passenger flows, fares, social welfare, delays, and disruptions. We will describe various data analysis methods for using incomplete datasets, for disaggregating aggregated datasets, and for modeling customer choice using revealed preferences datasets. Our analysis is based on three large-scale Big Data sets which are all available publically. We will describe applications of statistical methods including quantile regression, linear regression, multinomial logit models and mixed-logit models. Additionally, we will present various techniques that we used for data cleaning and preprocessing. We will end the presentation by providing several examples of practically useful insights drawn through these big data analyses. Presenter CV: Vikrant Vaze is an Assistant Professor in the Thayer School of Engineering at Dartmouth College. His research interests include applications of optimization, game-theory, machine learning and statistics to multi-agent complex systems, such as transportation and healthcare. He holds MS in Transportation, MS in Operations Research, and PhD in Systems, all from MIT. He also holds BS in Civil Engineering from IIT Mumbai. He has received several awards and honors including the inclusion in Who’s Who of America, Pikarsky Best PhD Dissertation Award from CUTC, the Best Paper Award from USA/Europe ATM R&D Seminar, Anna Valicek Best Paper Honorable Mention Award from AGIFORS, Graduate Research Award from the US Department of Transportation, MIT Presidential Fellowship, UPS Merit-based Fellowship at MIT, and President of India Gold Medal at IIT. Prior to joining Dartmouth, he has worked as an Algorithmic Trading Associate at Lehman Brothers Inc., as a Post-doctoral Associate at MIT and as a Research Scientist at Philips Research North America.
讲座时间:2016年6月7日14:00-15:30
讲座地点:address
讲座内容:Abstract: This talk will focus on data-driven and statistical modeling techniques using large real-world datasets. We will focus on three different case studies in the context of the US air transportation system. The US National Aviation System (NAS) is a complex system with multiple, interacting agents including airlines, passengers, and regulators, each with somewhat different objectives and incentives. These interactions determine the state of the system in terms of passenger flows, fares, social welfare, delays, and disruptions. We will describe various data analysis methods for using incomplete datasets, for disaggregating aggregated datasets, and for modeling customer choice using revealed preferences datasets. Our analysis is based on three large-scale Big Data sets which are all available publically. We will describe applications of statistical methods including quantile regression, linear regression, multinomial logit models and mixed-logit models. Additionally, we will present various techniques that we used for data cleaning and preprocessing. We will end the presentation by providing several examples of practically useful insights drawn through these big data analyses. Presenter CV: Vikrant Vaze is an Assistant Professor in the Thayer School of Engineering at Dartmouth College. His research interests include applications of optimization, game-theory, machine learning and statistics to multi-agent complex systems, such as transportation and healthcare. He holds MS in Transportation, MS in Operations Research, and PhD in Systems, all from MIT. He also holds BS in Civil Engineering from IIT Mumbai. He has received several awards and honors including the inclusion in Who’s Who of America, Pikarsky Best PhD Dissertation Award from CUTC, the Best Paper Award from USA/Europe ATM R&D Seminar, Anna Valicek Best Paper Honorable Mention Award from AGIFORS, Graduate Research Award from the US Department of Transportation, MIT Presidential Fellowship, UPS Merit-based Fellowship at MIT, and President of India Gold Medal at IIT. Prior to joining Dartmouth, he has worked as an Algorithmic Trading Associate at Lehman Brothers Inc., as a Post-doctoral Associate at MIT and as a Research Scientist at Philips Research North America.