Estimating Taxi Fares Using Machine Learning

Pierre Pascal Bian Theke

Master's thesis

Abstract:

This master's thesis focuses on the use of Artificial Intelligence (AI) algorithms for estimating taxi fares. By estimating fares in different zones of a city using these algorithms, the synchronization of the taxi fleet can be improved, which in turn would reduce waiting times. The thesis specifically analyzes the application of the Random Forest machine learning approach for this purpose. A literature review is presented on the development of AI in market analysis and price estimation. It emphasizes the shift towards machine learning for comprehensive market surveys. The process of developing accurate predictive models involves handling complex datasets and avoiding overfitting. To achieve this, the methodology includes configuring the model and justifying its architecture. Implementing the model involves pre- TBU in Zlín, Faculty of Applied Informatics 6 processing the data, training and validating it, and analyzing its performance in different scenarios. The thesis concludes by critically evaluating the accuracy of the Random Forest model and its interpretability, as well as its effectiveness in estimating fares. It highlights the ability of AI to change pricing strategies in the taxi industry.

Abstract:

This master's thesis focuses on the use of Artificial Intelligence (AI) algorithms for estimating taxi fares. By estimating fares in different zones of a city using these algorithms, the synchronization of the taxi fleet can be improved, which in turn would reduce waiting times. The thesis specifically analyzes the application of the Random Forest machine learning approach for this purpose. A literature review is presented on the development of AI in market analysis and price estimation. It emphasizes the shift towards machine learning for comprehensive market surveys. The process of developing accurate predictive models involves handling complex datasets and avoiding overfitting. To achieve this, the methodology includes configuring the model and justifying its architecture. Implementing the model involves pre- TBU in Zlín, Faculty of Applied Informatics 6 processing the data, training and validating it, and analyzing its performance in different scenarios. The thesis concludes by critically evaluating the accuracy of the Random Forest model and its interpretability, as well as its effectiveness in estimating fares. It highlights the ability of AI to change pricing strategies in the taxi industry.

Keywords

Artificial Intelligence, Machine Learning, Random Forest Algorithm, Taxi Fare Prediction, Market Analysis, Price Estimation, Data Acquisition, Model Configuration, Hyperparameter Selection, Data Pre-processing, Fare Estimation Effectiveness, Taxi Industry Revolution.

Language used: English

Date on which the thesis was submitted / produced: 13. 5. 2024

Thesis defence

Tomas Bata University in Zlín

Faculty of Applied Informatics

Master programme / field:
Engineering Informatics / Information Technologies