1999INFRASTRUCTURE PLANNING REVIEWOpen access

A Comparative Study on the Estimation Characteristics of Mixed Logit Model and Probit Model

Tetsuo Shimizu, Tetsuo Yai

Open full text 4 citations

Abstract

The aim of this study is to compare the estimation characteristics of mixed logit model and probit model with structured covariance for railway route choice. We improve the formulation for mixed logit model in order to be more theoretical formula. We analyze the case of 3 alternatives and it is suggested that probit model is prior to get more stable parameters rather than mixed logit model.In the further study, method for setting choice set may be obviously important issue.

Open-access reader

About this research paper

What this paper is about

The aim of this study is to compare the estimation characteristics of mixed logit model and probit model with structured covariance for railway route choice. We improve the formulation for mixed logit model in order to be more theoretical formula. We analyze the case of 3 alternatives and it is suggested that probit model is prior to get more stable parameters rather than mixed logit model.In the further study, method for setting choice set may be obviously important issue.

Why it matters

OpenAlex reports 4 citations for this work. Citation counts describe recorded attention and do not establish research quality.

Key contribution

A contribution statement is not available in the OpenAlex record.

Method / approach

Method details are not available in the OpenAlex metadata.

Main findings

Findings are not separately available in the OpenAlex metadata.

Limitations

Limitations are not available in the OpenAlex metadata.

Applications

Application details are not available in the OpenAlex metadata.

Available abstract

The aim of this study is to compare the estimation characteristics of mixed logit model and probit model with structured covariance for railway route choice. We improve the formulation for mixed logit model in order to be more theoretical formula. We analyze the case of 3 alternatives and it is suggested that probit model is prior to get more stable parameters rather than mixed logit model.In the further study, method for setting choice set may be obviously important issue.

Key concepts: Ordered probit, Mixed logit, Probit, Multinomial probit, Logit, Econometrics, Probit model, Logistic regression

Related papers

Back to paper searchBrowse research topicsOriginal source
A Comparative Study on the Estimation Characteristics of Mixed Logit Model and Probit Model — Research Paper | ScholarLens