This repository contains a deep learning-based classification model for predicting breast cancer using a neural network. The model is trained on the Wisconsin Breast Cancer dataset and can classify whether a tumor is malignant or benign based on various features.

• Overview
• Features
• Installation
• Usage
• Model
• Results
• Streamlit Application
• Contributing
• License

Breast cancer is one of the most common cancers among women. Early detection through classification can significantly improve treatment outcomes. This project uses a neural network to classify breast cancer tumors as either malignant or benign based on several features extracted from cell nuclei present in digitized images of a fine needle aspirate (FNA) of a breast mass.

• Deep Learning Model: A neural network trained for high accuracy classification.
• Preprocessing: Data is scaled using a pre-trained scaler to improve model performance.
• Streamlit Web App: An interactive web application built with Streamlit to allow users to input features and get predictions.

To run this project locally, follow these steps:

1. Clone the repository:

   git clone https://github.com/alihassanml/Breast-Cancer-Classification.git
   cd Breast-Cancer-Classification

2. Install the required dependencies:

   pip install -r requirements.txt

3. Make sure the following files are in the project directory:

   • model.h5: The pre-trained neural network model.
   • scalar.pkl: The scaler used for feature scaling.

You can run the Streamlit app to make predictions using the trained model:

streamlit run app.py

The app requires the following features to make a prediction:

• Mean Radius
• Mean Texture
• Mean Perimeter
• Mean Area
• Mean Smoothness
• Mean Compactness
• Mean Concavity
• Mean Concave Points
• Mean Symmetry
• Mean Fractal Dimension
• Radius Error
• Texture Error
• Perimeter Error
• Area Error
• Smoothness Error
• Compactness Error
• Concavity Error
• Concave Points Error
• Symmetry Error
• Fractal Dimension Error
• Worst Radius
• Worst Texture
• Worst Perimeter
• Worst Area
• Worst Smoothness
• Worst Compactness
• Worst Concavity
• Worst Concave Points
• Worst Symmetry
• Worst Fractal Dimension

The model outputs a prediction indicating whether the tumor is likely to be malignant or benign.

The neural network model is built using TensorFlow and Keras. It has been trained on the Wisconsin Breast Cancer dataset, achieving an accuracy of 98.40% on the test set.

[[ 66   1]
 [  2 119]]

              precision    recall  f1-score   support

           0       0.97      0.99      0.98        67
           1       0.99      0.98      0.99       121

    accuracy                           0.98       188
   macro avg       0.98      0.98      0.98       188
weighted avg       0.98      0.98      0.98       188

The Streamlit application provides a user-friendly interface to input the necessary features and receive a prediction. The app is designed for easy deployment and use.

Contributions are welcome! If you have any suggestions, feel free to open an issue or create a pull request.

This project is licensed under the MIT License - see the LICENSE file for details.