Data complexity is constantly increasing, in more and more cases data is generated as multivariate time series. While classical statistics offer solutions for many well known problems, certain systems and their internal dependency structures can not be analyzed using these classical approaches. J.O Ramsay and B.W Silverman introduced the framework of functional data analysis in order to model systems, which include multivariate time series. We will use a real world problem as a test environment and analyze the potential of functional data analysis. This will be done by extending the known framework and develop functional binary regression models. As part of the challenge two data processing methods called dynamic time warping and spline approximation are used to construct suitable data objects. We further show explorative possibilities to analyze functional data. Lastly we apply the derived theory to data from advanced driver assistance systems.     «

Data complexity is constantly increasing, in more and more cases data is generated as multivariate time series. While classical statistics offer solutions for many well known problems, certain systems and their internal dependency structures can not be analyzed using these classical approaches. J.O Ramsay and B.W Silverman introduced the framework of functional data analysis in order to model systems, which include multivariate time series. We will use a real world problem as a test environment...     »