台灣地區汽車客運路線成本歸類之研究

Abstract

本研究針對各汽車客運業者提送交通部審核受補貼路線，探討汽車客運路線與地區特性影響。 首先對於汽車客運路線不適當資料作調整，將原1107條汽車客運路線縮減為636條汽車客運路線，其中最高的六項成本為行車人員薪資、燃料費用、車輛折舊費用、業務員工薪資、管理員工薪資及修車員工薪資，佔路線每車公里總成本之83.25%的成本支出。 利用因子分析將原成本項目萃取出少數具有代表性的「區位」、「企劃」、「路線」、「車輛」、「規模」五項共同因素，進行汽車客運路線群落分群，並經由正典鑑別分析法，選擇汽車客運路線群落之鑑別力及配適度指標皆優良的「9群」，並建立鑑別函數，可為新汽車客運路線歸類。

The research is about examining and verifying subsidized routes by Ministry of Transportation and Communication from bus industry, and discussing about effect between bus routes and regions. At first, Taking modify about unsuitable information of bus routes, and cutting originally 1107 to 636 bus routes. There are six items are very high: bus driver salaries, fuel expense, vehicle depreciation expense, business personnel salaries, managerial personnel salaries and maintenance personnel salaries. The above distributions are 83.25% of total cost per kilometer a bus. By means of Factor Analysis, I select 5 common elements from originally cost items: regions, plans, vehicles, routes, and scales, and part bus route to clusters. In addition to what concluded above, I make use of Canonical Discriminate Analysis to find out “9 clusters” which are excellent about bus routes Discriminate Power and Good of Fitness. Setting up Discriminate Function to classify new bus routes.