Project working working with the NYC Taxi & Limousine Commission (TLC) to build a model that predicts taxi fares ahead of rides for an app they were building and to build a model that predicts whether or not customers would leave a tip to generate more revenue for taxi drivers.
How can we predict taxi fares and whether a customer will leave a tip ahead of rides?
Build, evaluate, and choose a predictive model that predicts taxi fares ahead of rides and another model that predicts whether a customer will leave a tip.
- Time Period: 2017
- Rows: 22,699
- Columns: 18
- Import necessary libraries/modules
- Perform initial EDA
- Conduct data cleaning
- Nulls
- Duplicates
- Outliers
- Perform full EDA
- Data wrangling and feature engineering
- Conduct A/B test to determine whether there is a relationship between payment type and fare amount
- Build/evaluate Multiple Linear Regression model to predict taxi fares ahead of rides
- Build/evaluate/choose between Random Forest and XGBoost models to predict whether a customer will leave a tip
- Provide final recommendations/next steps