A commuter who resides in Tokyo Urban Area has of ten two or more routes between his home and work place. When a commuter reached at transfer station, then there are two available route to his destination' one is shorter time and higher cost, another is longer time and lower cost, which route will he choose? As the results of survey and analysis, to save one minute, a commuter will pay at most 4 yen, and average payment is 1.5 yen. In assigning commuters to rapid transit facilities, the minimum-time-path method is used generally, but according to the research a commuter does not always take the minimum-time-path, and transfer the nearest transfer station from his work place. To analyse the route selection, the method of multivariate statistical analysis is introduced. As the principal variables which affect route sclection, four parameters are taken. These are cost, time, walking time, numbers of transfer. The value of these parameter are estimated about the altunative routr as well as using route. Commuters who use chuo line and are available Seibu-Shinjuku line are classified into π_1 group and commuters who use Seibu-Shinjuku line and are available chuo line are classified into π_2 group. It is tried to establish linear function. which are such that they distinguish most successfully in a certain sense between these two group. The method used for this purpose is that called discrininant analysis.
抄録全体を表示