| | | The objective of this project is to understand the issues affecting trams and buses in the TPG network by linking observed performance problems to spatial, temporal, and infrastructural factors, and by capturing how these effects vary over time and across network segments (e.g., peak hours impacting some segments but not others).
Methodologically, the project develops a latent variable model in which a segment�time�specific problematicness factor is inferred from multiple indicators through measurement equations, and then explained and predicted using structural equations that incorporate spatial attributes and time-varying factors. The focus is on dynamic estimation and prediction of problematic segments using real data, without relying on ad hoc thresholds on speed or delay.
Student should have followed Mathematical Modeling of Behavior (MATH-463) or have an equivalent statistical background. |