Modeling intracity tours with work purpose using weighted multiple regressions

In this paper by using the aggregate socioeconomic features of each zone and considering the land use prosperities, the aggregate zonal tours were modeled. For that, classic 4-step modeling database was utilized in which less data collection cost needs. Due heteroscedasticity , appling weighted least square (WLS) method led to multiple weighted regressions by two exogenous independent variables; population and number of employers in zone. Heteroscedasticity affect the efficiency of regression. In comparing of two model, the F test of WLS method growth 27 percent and the amount t students for population reduced from 2.79 to 2.27 that in negligible. On the other hand, the amount of t students for number of employers increases from 2.62 to 3.40 (30 percent). Constant coefficient is negative but in comparison with the maximum number of observed tours (7202 tours) is 0.069 percent and the average number of observed tours (1386 tours) is 0.36 percent that is negligible and the R2 goodness of fit index is 0.927 and acceptable.

Mots clés

Activity-based Model; Tour; Weighted Multiple Regressions; Qazvin.