Review of Classification Medical Images Using Ensemble Learning

Abstract

Machine learning and deep learning play an important role today in the field of classification and prediction of diseases, particularly through computer-assisted medical imaging and modern devices, which aid in early disease detection, lowering the actual risk, and assisting doctors in making a final diagnosis decision. Disease classification using ensemble learning techniques was developed to overcome the classification problem using a convolutional neural network (CNN). Although CNN produces good results, it requires a large dataset, limiting its use in classifying medical images due to its limited data availability and privacy. The research and literature have revealed that traditional machine learning algorithms perform poorly when trained using unbalanced datasets. This research aims to shed light on the use of ensemble learning in medical diagnosis, presents ensemble learning techniques used in classification, reviews previous research and work in which ensemble techniques were used, and proposes ensemble learning as a way to increase the accuracy and efficiency of classification systems by collecting the results of multiple classifiers and outputting the most voted results. These techniques improve the performance of a single model by using multiple models and combining their predictions.