📅 2024-09-05 — Session: Refinement of Nested Logit Model for Bus Choices
🕒 22:25–23:05
🏷️ Labels: Nested Logit Model, Probability, Transportation, Correlation Parameter, Logsum
📂 Project: Dev
⭐ Priority: MEDIUM
Session Goal
The session aimed to refine the nested logit model used to calculate the probability of travelers choosing between Express and Local bus options, focusing on the Red and Blue Express buses.
Key Activities
- Detailed exploration of nested logit model structure, including calculation of inclusive values and probabilities.
- Adjustment of correlation parameters to refine model accuracy.
- Analysis of discrepancies in expected versus actual probabilities for choosing an Express bus.
- Reassessment of logsum values and scaling factors to address potential structural issues in the model.
Achievements
- Identified a significant discrepancy in the probability of choosing an Express bus, with actual values exceeding expected targets.
- Highlighted the need for further refinement in the correlation parameter and scaling factor applications.
Pending Tasks
- Further refinement of the correlation parameter and scaling factors to align model outputs with expected probabilities.
- Additional analysis to clarify potential structural misunderstandings in the nested logit model setup.