Mushroom Classification Project

Overview

This project focuses on the classification of mushrooms into edible and poisonous categories based on various features. The classification model is built using the Flask web framework and a Random Forest Classifier.

Project Structure

The project is organized as follows:

• Mushroom_classifier:
  • data:
    • new_data.csv: Dataset containing preprocessed and imputed mushroom data.
  • inputs_preprocessing:
    • preprocessing.py: Module for preprocessing input data.
  • models:
    • trained_model.pickle: Pickle file containing the trained Random Forest model.
  • notebooks:
    • MushroomClassifier.ipynb: Jupyter Notebook containing the model training code.
  • static:
    • style.css: CSS file for styling of html.
  • templates:
    • index.html: HTML file for user input form.
    • result.html: HTML file to display the prediction result.
• .gitignore: File specifying files and directories to be ignored by version control.
• app.py: Flask application file.
• README.md: Project documentation.
• requirements.txt: File specifying Python dependencies.

How to Run

1. Install dependencies using pip install -r requirements.txt.
2. Run the Flask application: python app.py.
3. Access the application in your web browser at http://localhost:5000.

Usage

1. Visit the home page at http://localhost:5000.
2. Fill in the mushroom features in the form.
3. Click "Predict" to get the classification result.

Model Training

The model was trained using the Jupyter Notebook MushroomClassifier.ipynb. The best parameters for the Random Forest Classifier were determined using Grid Search.

Credits

• Dataset Source: UCI Machine Learning Repository
• Flask: Flask Documentation
• Scikit-learn: Scikit-learn Documentation
• Autoimpute: Autoimpute Documentation