Title:
Estimation and Simulation of a Small-Scale Activity-Based Model Using a Mixed Logit Simulator
Responsable(s) :
Anne-Valerie Preto, Prunelle Vogler, Michel Bierlaire
Description :
Activity-based models (ABMs) describe the daily schedules of individuals and their interactions within households. Since complete daily schedules are not enumerable, estimation must rely on simulation rather than analytical enumeration. This project aims to reproduce a small ABM instance for one household using the UK National Travel Survey (NTS) data. The student will define key attributes (e.g., activity duration, timing, mode, joint participation) based on existing parameter estimates and integrate the different components of a simulator and estimator within a mixed logit framework. The goal is to generate realistic daily activity patterns and evaluate how well the simulated schedules match observed ones.
Collaboration with:
Type :
semester project
Pré-requis :
Knowledge of discrete choice modeling and Python programming. Recommended course: Mathematical modelling of behavior (MATH-463)
Submitted on :
November 13, 2025