Event Type
Research Presentation
Academic Department
Mathematics and Statistics
Start Date
25-4-2022 12:00 AM
End Date
25-4-2022 12:00 AM
Description
The goal of this research is to create a fare prediction model for ride hailing companies Uber and Lyft in the Greater Boston area. We create multiple linear regression models for the two companies and compare the difference in their pricing strategies.
Uber and Lyft have made commute reliable and convenient, especially for individuals who do not own personal vehicles. Regular consumers of these services often experience unusual price fluctuations for a given source and destination. Finding a model that accurately predicts fares can help consumers decide the best choice for commute.
To build our models, we use a sample dataset available in Kaggle for Uber and Lyft price pings collected in Boston, MA. The dataset contains 110,190 data points for UberX and UberXL, and 102,470 data points for Lyft and LyftXL. We dropped rows with information regarding any other type of Uber/Lyft.
Predicting Uber and Lyft Fares Using Linear Regression
The goal of this research is to create a fare prediction model for ride hailing companies Uber and Lyft in the Greater Boston area. We create multiple linear regression models for the two companies and compare the difference in their pricing strategies.
Uber and Lyft have made commute reliable and convenient, especially for individuals who do not own personal vehicles. Regular consumers of these services often experience unusual price fluctuations for a given source and destination. Finding a model that accurately predicts fares can help consumers decide the best choice for commute.
To build our models, we use a sample dataset available in Kaggle for Uber and Lyft price pings collected in Boston, MA. The dataset contains 110,190 data points for UberX and UberXL, and 102,470 data points for Lyft and LyftXL. We dropped rows with information regarding any other type of Uber/Lyft.
Comments
Under the direction of by Dr. Giancarlo Schrementi.