New York Taxi Fare Prediction: A Machine Learning Practice Integrating Peak Hour Analysis and Inflation Data

This project addresses the uncertainty of New York taxi fares by using a random forest model that integrates peak hour analysis and inflation data from 2016 to 2025. It achieves a precise prediction with a root mean square error (RMSE) of only $1.79, providing a reliable reference for travelers, the taxi industry, and urban planning.

As an international metropolis, New York relies heavily on taxis as an important means of transportation. However, peak-hour congestion, late-night surcharges, and long-term inflation make fares difficult to predict, causing inconvenience to passengers. This project aims to address this practical need through machine learning technology.

1. Random Forest Algorithm: This algorithm was chosen for its strong ability to handle high-dimensional data, reduce overfitting, and its advantages such as feature importance analysis and nonlinear modeling; 2. Peak Hour Module: Identifies travel time types, estimates congestion coefficients, and converts them into features; 3. Inflation Tracking Mechanism: Incorporates base rates, per-mile prices, surcharge changes, and CPI inflation coefficients from 2016 to 2025 to ensure predictions reflect current price levels.

The system provides an intuitive interactive interface, and prediction only requires three steps: 1. Enter the start and end locations (supports address text and map point selection); 2. Select the travel time (automatically determines peak hours); 3. Obtain the prediction result and detailed fare breakdown (base fare, mileage fee, surcharge, etc.).

The project's core indicator RMSE is $1.79. For New York taxi trips with an average fare of $15-$30, the accuracy reaches over 90%. This accuracy can help passengers budget accurately, tourists avoid unreasonable charges, and drivers/platforms optimize pricing strategies.

• Personal Travel: Compare costs of different travel modes, choose optimal times, and make budgets; - Business Decisions: Optimize dynamic pricing, reduce fare disputes, and analyze supply and demand; - Urban Planning: Identify congestion hotspots, evaluate weak links in public transportation, and support industry policies.

Future optimization directions for the project: 1. Integrate real-time traffic data to improve dynamic prediction; 2. Incorporate weather factors into modeling; 3. Expand multi-mode travel comparison; 4. Develop a mobile application for convenient on-the-go queries.