Master’s student Walaa Abdul-Latif defended her thesis titled “A Technique for Classifying Fundus Diseases Using Machine Learning and Deep Learning”, submitted to the College of Computer Science and Information Technology / Department of Computer Science, under the supervision of Assistant Professor Dr. Ali Mohsen Mohammed.

The thesis aimed to design and develop two models for classifying fundus diseases using modern techniques that combine deep learning and machine learning. The first model focused on applying a hybrid technique based on Convolutional Neural Networks (CNNs) to extract features from images and classify them using machine learning algorithms. In contrast, the second model relied on Local Descriptor Pattern Variance (LDPV) for feature description and selected the best features using Neighbor Component Analysis (NCA).

Among the key objectives of the study were improving the accuracy and speed of fundus disease diagnosis, reducing reliance on manual diagnosis—which requires significant time and effort—and addressing challenges related to multi-class image classification by implementing innovative techniques to enhance classification efficiency and accuracy.