Eye Age and Health Classification Project (ML Hackathon)

Overview

The Eye Age and Health Classification project focuses on predicting the age of patients and classifying the health status of their eyes as normal or indicating a disease. The project utilizes transfer learning with a pre-trained ResNet model and incorporates additional patient information for enhanced prediction accuracy.

Getting Started

Prerequisites

• Python 3.7
• Conda (optional but recommended for environment management)

Installation

1. Clone the repository:

   git clone https://github.com/HasanOJ/eye-age-regression.git
   cd eye-age-regression

2. Create a virtual environment (optional but recommended):

   conda env create -f environment.yaml
   conda activate MLhackathon

Usage

Run the main.py script from the terminal with the following command-line arguments:

python main.py -d /path/to/data -t left -c /path/to/checkpoint --train -e 10

Command-line Arguments:

• -d or --data_path: Path to the data directory (default: "data").
• -t or --tag: Tag for the model (default: "left").
• -c or --checkpoint: Path to the checkpoint file (optional).
• --train: Flag to indicate training (default: False).
• -e or --epochs: Number of epochs for training (default: 10).

Example command to test the model for right tag:

python main.py -d /path/to/data -t right

Dataset

The dataset contains information about patients' fundus images, including their age, left and right eye images, and a binary label indicating whether the eyes are normal (healthy) or indicate a disease.