Machine Learning in Economics: Case Study of Tipping Behavior in Taxicab Industry

Date of Award

2019

Document Type

Thesis

Degree Name

Bachelor of Arts

Department

Economics

First Advisor

Qi Ge

Abstract

This paper presents a conceptual framework for thinking about machine learning in econometrics, introduces machine learning techniques like LASSO, decision tree, and gradient boosted decision trees, and reviews articles from three categories of machine learning applications in economics. Our empirical setting employs a high-frequency dataset on taxicab rides in New York City to predict passengers’ tipping behavior using several learning algorithms. We find machine learning algorithms to bring clear improvement in prediction performance. However, the practical meaning of the improvement is clouded in light of low prediction performances across the board.

This document is currently not available here.

Share

COinS