An Approach of Hybrid models based on Fuzzy reasoning and Logit models
Kaori Mizutani, Takamasa AKIYAAMA · INFRASTRUCTURE PLANNING REVIEW · 2001
Hybrid models based on the random utility models and the soft computing techniques have been proposed in order to describe various aspects of travel behaviour effectively. The hybrid models are identically systemized considering with the forms of the model formulations in the paper. Two hybrid models called fuzzy logit models are constructed and tested. One is a logit model with a fuzzy reasoning-based utility function that can describe human decisions. The utility function with highly non-linear relations illustrates the diminishing marginal utility. As for the other model, bothdeterministic and probabilistic travel behaviour are described using the fuzzy reasoning and the logit model, independently.