Traffic Level Prediction

Project Overview

This project aims to predict traffic levels based on historical data using a machine learning model. It utilizes a Logistic Regression model trained on traffic data to classify whether there will be heavy traffic or not on a given date and junction. The prediction is made through a user-friendly Streamlit web application.

The model considers features such as hour, day, month, day of the week, whether it's a weekend, rush hour, and holidays to make accurate predictions.

Folder Structure

Traffic_Level_Prediction/
├── requirements.txt          # Python dependencies
├── data/
│   ├── traffic_data.csv      # Raw traffic data
│   └── cleaned_traffic_data.csv  # Processed data after cleaning and feature engineering
└── src/
    ├── clean_data.py         # Script to clean and preprocess the data
    ├── feature_eng.py        # Feature engineering functions
    ├── Logistic_Regression_model.py  # Model training script
    ├── main.py               # Streamlit app for predictions
    ├── __pycache__/          # Python cache files
    └── models/
        └── Logistic_Regression_model.pkl  # Trained model file

Libraries Used

The project relies on the following key Python libraries:

• Pandas: For data manipulation and analysis.
• Scikit-learn: For machine learning algorithms, specifically Logistic Regression.
• Streamlit: For building the interactive web application.
• Joblib: For saving and loading the trained model.

Installation

1. Clone the repository:

   git clone https://github.com/yourusername/Traffic_Level_Prediction.git
   cd Traffic_Level_Prediction
   

2. Install the required dependencies:

   pip install -r requirements.txt
   

Usage

1. Run the data cleaning and feature engineering:

   python src/clean_data.py
   

2. Enter the src folder:

   cd src
   

3. Train the model:

   python Logistic_Regression_model.py
   

4. Launch the Streamlit app:

   streamlit run main.py
   

5. Open your browser and go to the provided URL to use the prediction app.

How It Works

• The app takes a date input and junction selection.
• It extracts features like hour, day, month, day of week from the date.
• The trained Logistic Regression model predicts if there will be heavy traffic (more than 20 vehicles) or not.
• The result is displayed to the user. d:\work\GitHub\Traffic_Level_Prediction\README.md