MSCI446 Final Project

Uber Rideshare Predictive Pricing 🚗

Final Notebook: https://colab.research.google.com/drive/1aDEnMt3r6saN5i83sY8X2I3G7I7ZD4DM

Ella Kennedy, Manjary Muruganandan, Kiranjit Nagra, Aryan Chopra, Achchala Deepan

Steps to get started:

1. run pip install -r requirements.txt to install dependencies in requirements.txt
2. Download the rideshare_kaggle.csv, cleaned_data.csv, and the data_feature_engineering.csv file from the zipped data file uploaded to Learn
3. Add the rideshare_kaggle.csv (which is uploaded on Learn) to this folder for the data preprocessing steps in uber_predictive_pricing.ipynb to the data folder in this repository
4. Add the cleaned_data.csv file (which is generated with the data preprocessing steps and also uploaded on Learn) into the data folder for the feature engineering steps in uber_predictive_pricing.ipynb
5. Add the data_feature_engineering.csv (which is generated at the end of the feature engineering steps and also uploaded on Learn) into the data folder, to be used the rest of the project (test/train split, modelling) in uber_predictive_pricing.ipynb

Aim

Empirical Evaluation: collecting data and proposing ML solutions to an area of real-life application

Business Problem and Motivation

The general focus of this project is to create an optimal predictive pricing model to forecast Uber rideshare services, with a goal of enhancing current dynamic pricing strategies. This stems from the observation that traditional pricing models often do not fully account for the plethora of factors influencing ride costs, which can lead to lessened user satisfaction. By leveraging the comprehensive Boston, MA Uber and Lyft Kaggle dataset, our project seeks to incorporate a wider range of variables into the pricing algorithm.