COMPARISON OF NON-COMPENSATORY MODELS OF DRIVER'S CHOICE ON DYNAMIC PARK AND RIDE

Abstract

Non-compensatory models of driver's choice on dynamic park and ride are developed and examined on the predictability in this study. One of the data mining tools, C4.5, is used to develop decision tree and production rules of driver's choice. The generated decision tree and production rules are compared with the semi-ordered lexicographic model developed in the preceding study. Compared are the similarity of the estimated decision making structures and the distribution of the segments not correctly represented by the models as well as goodness-of-fit and hit ratio. The comparison shows that the semi-ordered lexicographic model has a higher goodness-of-fit and hit ratio than models with data mining tools. The results also suggest that the models developed in this study represent different decision rules from, but have similar distributions of the segments not represented by the models with the semi-ordered lexicographic model. Empirically, the consistent results by models with data mining tools and the semi-ordered lexicographic model suggest that the parking congestion level in CBD has a significant effect on the choice behavior.

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 — 8 citations exist upstream.

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