Comprehensive Machine Learning Pipeline on Heart Disease UCI Dataset

This project provides a complete Machine Learning pipeline for analyzing, predicting, and visualizing heart disease risks using the UCI Heart Disease dataset. It covers data preprocessing, dimensionality reduction (PCA), supervised & unsupervised models, hyperparameter tuning, and optional web deployment using Streamlit & Ngrok.

• Project Objectives
• Dataset
• Tools & Libraries
• Project Structure
• How to Run
  • Clone the Repository
  • Create the Virtual Environment
  • Activate the Virtual Environment
  • Install Dependencies
  • Run Jupyter Notebooks
  • Run the Streamlit Web App
  • Deploy using Ngrok
• Pipeline Workflow
• Results & Deliverables

• Perform Data Cleaning & Preprocessing (missing values, encoding, scaling).
• Apply Dimensionality Reduction using PCA.
• Implement Feature Selection using Random Forest, RFE, Chi-Square.
• Train Supervised Models:
  • Logistic Regression
  • Decision Tree
  • Random Forest
  • Support Vector Machine (SVM)
• Apply Unsupervised Learning:
  • K-Means Clustering
  • Hierarchical Clustering
• Optimize models using GridSearchCV & RandomizedSearchCV.
• Deploy a Streamlit Web App and use Ngrok for public access.

• Name: Heart Disease UCI Dataset
• Description: Predict the presence or absence of heart disease based on clinical parameters.

• Language: python
• Libraries:
  • pandas, numpy – Data Handling
  • matplotlib , Seaborn – Visualization
  • sklearn – Machine Learning Models & PCA
  • joblib - Save Model as .plk
  • streamlit – Interactive Web App
  • ngrok – Public URL for Deployment

Heart_Disease_Project/
│── data/
│   ├── heart_disease.csv
│── notebooks/
│   ├── 00_data_collecting.ipynb
│   ├── 01_data_preprocessing.ipynb
│   ├── 02_pca_analysis.ipynb
│   ├── 03_feature_selection.ipynb
│   ├── 04_supervised_learning.ipynb
│   ├── 05_unsupervised_learning.ipynb
│   ├── 06_hyperparameter_tuning.ipynb
│── models/
│   ├── final_model.pkl
│── ui/
│   ├── app.py (Streamlit UI)
│── deployment/
│   ├── ngrok_setup.txt
│── results/
│   ├── evaluation_metrics.txt
│── requirements.txt
│── README.md
│── .gitignore

git clone https://github.com/basmala-ayman/Heart-Disease.git
cd Heart-Disease

python3 -m venv venv

# On macOS/Linux:
source venv/bin/activate

# On Windows:
venv\Scripts\activate

pip install -r requirements.txt

jupyter notebook

streamlit run ui/app.py

Read instructions in deployment/ngrok_setup.txt.

1. Data Preprocessing & Cleaning – Handle missing values, encoding, scaling
2. PCA Analysis – Dimensionality Reduction
3. Feature Selection – Random Forest, RFE, Chi-Square
4. Model Training – Logistic Regression, Decision Tree, Random Forest, SVM
5. Evaluation – Accuracy, Precision, Recall, F1, ROC-AUC
6. Clustering – K-Means & Hierarchical Clustering
7. Hyperparameter Tuning – GridSearchCV, RandomizedSearchCV
8. Deployment (Bonus) – Streamlit & Ngrok

• Cleaned Dataset
• PCA & Feature Selection Results
• Trained Models with Evaluation Metrics
• Optimized Model Saved as .pkl
• Interactive Streamlit UI
• Ngrok Public Access Link