Breast Cancer Classification with Logistic Regression

This project demonstrates a beginner-friendly implementation of Logistic Regression to classify breast cancer tumors as benign or malignant, using the Breast Cancer Wisconsin dataset.

---

Objectives

• Preprocess and clean the dataset (remove NaNs, encode labels)
• Perform Exploratory Data Analysis (EDA) with visual insights
• Train and evaluate a Logistic Regression classifier
• Tune decision threshold and explain sigmoid output
• Save all visuals and results for GitHub display
• Present the work in a clean, job-ready format

---

Project Structure

breast-cancer-logistic-regression/ │ ├── data/ # Raw & cleaned dataset ├── images/ # Visuals (histograms, heatmaps, confusion matrix, ROC) ├── notebooks/ │ ├── 01_data_cleaning.ipynb │ ├── 02_eda.ipynb │ └── 03_logistic_model.ipynb ├── requirements.txt # Required libraries ├── LICENSE # MIT License └── README.md # This file

---

Dataset

• Source: Kaggle – Breast Cancer Wisconsin Dataset
• Shape: 569 samples, 32 features
• Target: diagnosis
  • 0 = Benign
  • 1 = Malignant

---

Tasks Completed

Task	Status
Data Cleaning & Preprocessing	✅
Exploratory Data Analysis (EDA)	✅
Logistic Regression Training	✅
Evaluation (Accuracy, F1, AUC, etc)	✅
Confusion Matrix + ROC Curve	✅
Threshold tuning	✅
All visuals saved in images/	✅

---

Model Evaluation

Metric	Score
Accuracy	93.86%
Precision	0.97
Recall	0.86
F1 Score	0.91
ROC-AUC Score	0.986

Confusion matrix and ROC curve visualizations are saved in the images folder.

---

Tools & Environment Used

• Jupyter Notebook – Interactive coding and documentation
• VS Code – Code editor used for development
• Git – Version control
• GitHub – Project hosting and portfolio building
• Kaggle – Dataset source
• Markdown – For clean documentation
• Python 3.10+ – Language used

---

Libraries Used

• Python 3
• Pandas, NumPy
• Matplotlib, Seaborn
• scikit-learn

---

How to Run This Project

# Clone this repository
git clone https://github.com/your-username/breast-cancer-logistic-regression.git
cd breast-cancer-logistic-regression

# Install required libraries
pip install -r requirements.txt

# Run notebooks
jupyter lab