Decision Tree approach to Identify Vision Disorder for lazy Eye

Abstract

Lazy Eye is a disorder of vision and is related to a neurological disorder, in which the brain is not able to receive correct input from one eye. This disease called Amblyopia. Two to five percent people are affected by this disorder. Cibis, Wang, and Van Eenwyk have developed an automatic system for sight screening, aimed for early finding the problem while the person is a child. These systems not able to achieve an accuracy of more than 78%. From AVVDA system two more features are used but achieved a low accuracy. We are using traditional machine learning technique, decision tree approach, to get more accuracy. We have used random forest WEKA architecture for the patient vision data set and for selecting the appropriate features we have used InfoGain Class. With the decision tree approach, our work showed good results, where the accuracy achieved is more than 90%.