Breast Cancer Prediction Application

Description

This web application predicts whether a breast tumor is benign or malignant using a pre-trained Support Vector Machine (SVM) model. The user-friendly interface allows inputting tumor features through a form to obtain an immediate diagnosis.

Features

• Predicts breast cancer type (benign or malignant) based on clinical features.
• Input data is normalized using a scaler for consistency.
• Implements and evaluates four different machine learning algorithms.
• Displays accuracy comparison as a bar chart for better insights.
• User-friendly web interface built with Flask.

Evaluation and Visualization

The model performance is compared using a bar chart, where the accuracies of the four algorithms are displayed:

• k-Nearest Neighbors (k-NN)
• Support Vector Machine (SVM)
• Logistic Regression
• Naive Bayes

SVM Model

The SVM model was trained on the Breast Cancer Wisconsin Dataset. The data was preprocessed, and a scaler was used to normalize input values. The svm_model.pkl file contains the saved model using the joblib library.

Prerequisites

Before starting, ensure the following are installed on your machine:

• Python 3.x
• Flask
• NumPy
• pandas
• scikit-learn
• joblib
• Matplotlib
• Seaborn

Installation

1. Clone this GitHub repository to your local machine:

   git clone https://github.com/AsmaeKarmouchi/Breast_cancer_prediction.git