Driver-vehicle-environment system characterization using statistical analyses
暂无分享,去创建一个
The paper shows that multiple correspondence analysis (MCA) is well suited to characterize the driver-vehicle-environment system. First, the complexity of this system is explained. Then partial models of this system are exposed followed by characterizing the methods. Among these methods, MCA has the required features to reveal what are the most relevant variables that could better characterize the system and to associate variable value tendencies with functioning modalities of the system. An example of MCA applied to characterize four driving situations is given.
[1] D. Kell,et al. Classification of pyrolysis mass spectra by fuzzy multivariate rule induction-comparison with regression, K-nearest neighbour, neural and decision-tree methods , 1997 .
[2] M. Hayhoe,et al. What controls attention in natural environments? , 2001, Vision Research.
[3] L. Evans. Risks older drivers face themselves and threats they pose to other road users. , 2000, International journal of epidemiology.