Abstract The travel times of various types of activities by a sample of Vancouver households are analyzed using a mutual information statistic. The basic objective of these analyses was to isolate those household factors influencing the consumption of travel. The advantage of the mutual information statistic over the normally used parametric statistics is that nominal variables such as stage in life cycle and marital status may be analyzed along with ordinal and ratio scale variables. The variables describing households are first analyzed and the information redundancy between the various variables is displayed through the use of a maximally weighted tree. Cluster analysis is used to develop clusters of households with similar travel time consumption profiles and the information statistic is used to identify the most important household descriptor variables. The method of analysis described in the paper is particularly useful for transport systems analysis where nominal and ordinal variables are frequently encountered.
Analyses of household travel activities by information statistics
Transportation Research Part A: General ; 15 , 2 ; 163-171
1980-02-25
9 pages
Article (Journal)
Electronic Resource
English
Making Household Microsimulation of Travel and Activities Accessible to Planners
British Library Conference Proceedings | 2005
|Making Household Microsimulation of Travel and Activities Accessible to Planners
Transportation Research Record | 2005
|Making Household Microsimulation of Travel and Activities Accessible to Planners
Online Contents | 2005
|