Breast Cancer Classification Using Machine Learning

This project utilizes machine learning techniques to classify breast tumors as malignant or benign based on the Breast Cancer Wisconsin (Diagnostic) Dataset. The goal is to develop a predictive model that can assist in the early detection of breast cancer.

About the Dataset

The Breast Cancer Wisconsin (Diagnostic) Dataset comprises 700 instances of breast cancer cases, each with nine numerical features. The target variable (Class) indicates whether the tumor is benign (2) or malignant (4).

Project Overview

The project involves the following steps:

1. Data Preprocessing:
   Handling missing values, encoding categorical variables, and scaling features to prepare the data for modeling.

2. Exploratory Data Analysis (EDA):
   Analyzing the distribution of features and their relationships with the target variable to gain insights into the data.

3. Model Development:
   To build predictive models, Implementing machine learning algorithms, including Decision Tree, KNN, SVM, Random Forest, NB, and Logistic Regression.

4. Model Evaluation:
   Assessing the performance of the models using metrics such as accuracy, precision, recall, F1-score, and ROC-AUC.

5. Feature Importance Analysis:
   Identifying the most significant features contributing to the model's predictions to understand the factors influencing breast cancer detection.

Results

The models achieved high accuracy in classifying tumors with the following performance metrics:

• Decision Tree:

  • Accuracy: 94%
  • Precision: 94%
  • Recall: 93%
  • F1-score: 94%

• K-nearest neighbors (KNN):

  • Accuracy: 90%
  • Precision: 92%
  • Recall: 89%
  • F1-score: 90%

• Support Vector Machine (SVM):

  • Accuracy: 91%
  • Precision: 93%
  • Recall: 89%
  • F1-score: 91%

• Random Forest:

  • Accuracy: 91%
  • Precision: 93%
  • Recall: 89%
  • F1-score: 91%

• Naïve Bayes (NB):

  • Accuracy: 91%
  • Precision: 93%
  • Recall: 89%
  • F1-score: 91%

• Logistic Regression:

  • Accuracy: 89%
  • Precision: 92%
  • Recall: 87%
  • F1-score: 89%

These results demonstrate the effectiveness of machine learning models in accurately detecting breast cancer based on tumor characteristics.

Usage

To replicate the analysis:

Google Colab: Link (File -> Save a Copy in Drive): Google Colab

Or

1. Clone the Repository:

   git clone https://github.com/DaveMatNat/breast-cancer-ml.git
   
   cd breast-cancer-ml

2. Install the Required Dependencies

   pip install -r requirements.txt

3. Run the Jupyter Notebook:

Launch Jupyter Notebook and open Breast_Cancer_ML.ipynb to explore the analysis and results.

---