TRANSFERABILIY AND UPDATING ANALYSIS OF MODE CHOICE MODEL BETWEEN TWO DEVELOPING COUNTRIES—PREDICTIVE PERFORMANCE POINT OF VIEW

Abstract

This paper presents the predictive performance analysis of transferring and updating a mode choice model estimated in Ho Chi Minh City to Phnom Penh. Difficulties in transferring a model due to different data collection method and simplified approaches to solve the problems are also presented. Four updating approaches associated with small sized sample of 200, 400 and 600 observations are applied in this study. The study shows that predictive performance improves with updating alternative specific constants and improves further with additional adjustment of scale parameter. Combined transfer estimator has proved to produce the most significant improvement in updating the model. In case of the existence of transfer bias, application of Bayesian updating is not recommended.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 7 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.