This project focuses on a comprehensive time series analysis of Chicago taxi trip data. By leveraging advanced modeling techniques and visualization tools, the project aims to uncover trends, analyze seasonality, and forecast taxi trip demand. The analysis includes implementing ARIMA and Prophet models, conducting stationarity tests, and creating an interactive Streamlit app for visualization and predictions.

The dataset contains historical Chicago taxi trip data, including trip dates and total trips.

Data is preprocessed to handle missing values and formatted for time series analysis.

Augmented Dickey-Fuller (ADF) Test: Determines if the data is stationary or requires differencing.

Kwiatkowski-Phillips-Schmidt-Shin (KPSS) Test: Validates the stationarity assumption.

Auto ARIMA: Automatically determines the optimal parameters (“p”, “d”, “q”).

Captures temporal patterns for short-term forecasting.

Includes seasonality adjustments to improve accuracy.

Robust forecasting model developed by Facebook.

Handles strong seasonality, holidays, and changepoints automatically.

Generates interpretable and reliable forecasts.

Plotly: Interactive and dynamic plots for trend analysis, model evaluation, and forecasting.

Includes raw data visualizations, yearly aggregations, and forecast overlays.

User-friendly and interactive web application.

Features include:

Hosted at: Streamlit Deployment

Trend Analysis:

Stationarity:

Model Performance:

Visualization:

Clone this repository:

Install dependencies:

Run the Streamlit app:

This project demonstrates the effective use of time series analysis techniques and interactive tools to understand and forecast taxi trip demand. By combining ARIMA and Prophet models with dynamic visualizations, the analysis provides valuable insights for decision-makers and stakeholders in the transportation industry.

Feel free to explore the code, results, and the interactive app. For any queries or suggestions, please reach out using the contact details below.