GTC Internship - Diabetes Prediction using Machine Learning

Repository Description

This repository contains an end-to-end machine learning project that predicts whether a patient is diabetic or non-diabetic based on diagnostic health measurements. The project covers the full ML workflow: exploratory data analysis (EDA), preprocessing, model building, evaluation, and deployment of a simple prediction function.

Project Workflow

Phase 1: Data Exploration

• Explore dataset distributions and class balance.
• Visualize relationships (e.g., glucose vs outcome, BMI vs outcome).
• Use charts and summary statistics to uncover hidden insights.

Phase 2: Data Preprocessing

• Standardize features for fair comparison across models.
• Split the dataset into training and testing sets.
• Ensure the model generalizes well to unseen data.

Phase 3: Model Building & Training

• Implement at least two ML models:

  • Logistic Regression
  • Support Vector Machine (SVM)
  • Random Forest (optional, for robustness)

• Apply hyperparameter tuning using GridSearchCV.

• Compare model performance based on accuracy and other metrics.

Phase 4: Prediction Engine

• Build a function that:

  • Accepts new patient data as input.
  • Returns an instant prediction → Diabetic or Non-Diabetic.

Repository Structure

diabetes-prediction-ml/
│-- data/                # Dataset (if shareable)
│-- notebooks/           # Jupyter notebooks for each phase
│-- src/                 # Source code (preprocessing, models, utils)
│-- results/             # Saved graphs, charts, and reports
|-- LICENSE              # MIT License
│-- README.md            # Project documentation

⚙️ Technologies Used

• Python 🐍
• Pandas & NumPy → Data manipulation
• Matplotlib & Seaborn → Data visualization
• Scikit-learn → ML models & evaluation

Evaluation Metrics

• Accuracy
• Precision
• Recall
• F1-Score
• Confusion Matrix

How to Run

1. Clone this repo:

   git clone https://github.com/abanoub-refaat/gtc-diabetes-prediction-model.git
   cd gtc-diabetes-prediction-model

2. Open the notebooks to explore each phase:

   jupyter notebook notebooks/

3. Run the prediction function (example in notebooks or src/predict.py).

Project Goals

• Understand and apply the end-to-end ML pipeline.
• Gain insights from EDA and data preprocessing.
• Compare models and improve performance with tuning.
• Build a working prediction tool for diabetes detection.

References & Learning Resources

• Scikit-learn Documentation
• EDA for Classification
• Why Standardize Data for ML?
• Hyperparameter Tuning with GridSearchCV

Author

Developed by Abanoub Refaat as part of the GTC Machine Learning Internship project.