Loading...

Spatial Transferability of Activity Choice and Time Allocation Joint Model

Nohekhan, Amir | 2016

862 Viewed
  1. Type of Document: M.Sc. Thesis
  2. Language: Farsi
  3. Document No: 49357 (09)
  4. University: Sharif University of Technology
  5. Department: Civil Engineering
  6. Advisor(s): Samimi, Amir
  7. Abstract:
  8. Spatial transferability of behavioral models can significantly decrease the costs and time required for transportation studies. The main aim in this study is exploring transferability of MDCEV joint model used for activity type and duration choices with implementing various transfer methods on different sample sizes. In order to achieve this goal, full information of comprehensive transportation studies conducted in Mashhad 1994 and Shiraz 1999 has been employed. Shiraz city is the region where the model is generated using its data and Mashhad is the region which the generated model is transferred to. Transfer methods studies in this study are Transfer scaling, Bayesian method, Combined transfer estimation and Joint context estimation. Different sample sizes are 1, 5, 10, 20, 30, 40 and 50 percents of Mashhad dataset. Results indicated that the model generated solely with Shiraz data is not transferable unless a transfer method is used to adjust the model. Various measures are used to examine the prediction ability of the transferred model. These measures have shown that Joint context estimation method is the best method and after that stands Combined transfer estimation
  9. Keywords:
  10. Multiple Discrete-Continuous Extreme Value Model (MDCEV) ; Transferability ; Bayesian Analysis ; Activity Type ; Activity Duration ; Model Spatial Transfer ; Transfer Scaling ; Combined Transfer Estimation ; Joint Context Estimation

 Digital Object List

 Bookmark