A comprehensive project on glaucoma prediction using machine learning. This repository includes data preprocessing, model training (Random Forest, Gradient Boosting), and evaluation metrics. The dataset is sourced from Kaggle, and techniques like SMOTE are applied for handling class imbalance.
Repository Description This repository contains a machine learning project focused on predicting the risk of glaucoma using clinical metrics and Optical Coherence Tomography (OCT) image data. Leveraging a dataset sourced from Kaggle, the project employs various algorithms, including Random Forest and Gradient Boosting, to analyze features and classify the likelihood of glaucoma in patients.
Key Features:
- Data Preprocessing: Includes handling of imbalanced data using SMOTE (Synthetic Minority Over-sampling Technique).
- Model Training: Implemented multiple classifiers to assess performance, including Random Forest and Gradient Boosting.
- Evaluation Metrics: Utilized accuracy, confusion matrix, and ROC-AUC scores to evaluate model performance.
- Feature Importance: Analyzed the importance of different clinical metrics in predicting glaucoma risk.
- Visualization: Provided visualizations for model evaluation, including confusion matrices and ROC curves.
This project aims to contribute to personalized health insights and future risk forecasting for glaucoma patients.
Technologies Used:
-> Python
-> Scikit-learn
-> Imbalanced-learn
-> Matplotlib
-> Seaborn
-> Google Colab