Premium
Modeling default behavior in the presence of information and its application to the route choice problem
Author(s) -
Lotan Tsippy,
Koutsopoulos Haris N.
Publication year - 1999
Publication title -
international journal of intelligent systems
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.291
H-Index - 87
eISSN - 1098-111X
pISSN - 0884-8173
DOI - 10.1002/(sici)1098-111x(199905)14:5<501::aid-int3>3.0.co;2-t
Subject(s) - computer science , discounting , perception , population , complete information , artificial intelligence , machine learning , psychology , mathematical economics , mathematics , demography , finance , neuroscience , sociology , economics
A new approach for modeling default behavior is presented and applied to route choice behavior in the presence of information. It is based on the assumption that drivers follow their usual behavior pattern and modify it only if the information received differs substantially from their existing perceptions and knowledge. Default behavior is linked to the interactions between existing knowledge and new information. These interactions are modeled through measures of compatibility, and several measures are suggested and analyzed. The approximate‐reasoning model, which was earlier suggested for modeling discrete choices made in the presence of information, is adapted to handle default behavior. The model is briefly presented and its implementation and calibration are discussed. Default behavior is implemented by discounting existing knowledge when it does not agree with the new information, and by discounting the new information when it does not significantly differ from existing knowledge. A small case study is conducted using a driver simulator to collect data from two types of drivers: familiar and unfamiliar. The results obtained provide interesting insights on the choice behavior of the sample population, and support a default type of behavior among the familiar drivers. ©1999 John Wiley & Sons, Inc.