Project Title : Classification on Diabetic Retinopathy in an image set

This project builds a PR system with Diabetic Retinopathy image data. We used four classifiers for training and testing purposes.

Traditional classifiers –

• Quadratic classifier
• K Nearest neighbour

Linear classifiers –

• The Ho-Kashyap rule
• Fisher’s Discriminant Method

Dataset Description for The Project

This project used Diabetic Retinopathy DebrecenDataset. It contains features extracted from the Messidor image set to predict whether an image contains signs of diabetic retinopathy or not.

Data set Characteristics	Attribute Characteristics	Number of Instances	Number of Attributes	Missing Values	Associated Tasks
Multivariate	Integer, Real	1151	20	N/A	classification

instruction to run:

python diabetic_classification.py