A Comparative Analysis of K-Nearest Neighbours & Support Vector Machine for Classification of Iris African Dataset
Keywords: Machine learning, Classification algorithm, Feature extraction, Eyes, Disease detection.
TL;DR: Classification of An Iris African Dataset
Abstract: Early eye illness detection is a major issue; early eye disease identification is crucial to preventing future complications. Early identification is crucial in several vision-losing disorders such as cataracts, diabetic retinopathy, and diabetes mellitus cataract, which cause blindness in working people at younger ages. This study aims to develop an eye disease detection model. The model was created by using an African iris dataset from Kaggle, PCA, KNN, and SVM (Support Vector Machine). The results decided which algorithms classified myopia or hyperopia best. Evaluation metrics were used to evaluate the performance implementation. The SVM algorithm outperformed the other algorithms, achieving a classification testing accuracy with PCA of 71.6%. The study concluded that the proposed approach can be used to accurately classify eye diseases in African patients and highlights the importance of considering the specific population when developing models for classifying or detecting eye diseases.
Submission Category: Machine learning algorithms
Submission Number: 32
Loading