Comparative Analysis of the Identification and Categorization of the Malaria Parasite Employing Recent Amalgamated Machine Learning Methodologies
Tamal Kumar Kundu, Dinesh Kumar Anguraj, R. Nidhya, V. Maruthi Prasad · 2025
Malaria persists as a prominent global health obstacle, necessitating inventive solutions to achieve a precise and effective diagnosis. This research introduces a multifaceted approach to malaria parasite detection and classification through three distinct methodologies. The aims center on the cultivation of superior models for automated identification and categorization, as well as the investigation of innovative methodologies for adjusting hyper-parameters within the realm of deep learning and introducing a tuneable Squeeze-Net prediction model for enhanced performance. Throughout the research, we emphasize comprehensive data collection, preprocessing, data augmentation, Feature extraction, and Image classification to ensure the model's robustness and uniqueness. Here in this article, we compare the models’ performance based on the model's different parameters such as accuracy, precision, Recall factor etc. This research comparison articulates methodologies, elucidating contemporary strides in employing machine-learning strategies for malaria diagnosis through microscopic imagery. The paper systematically compares diverse approaches employed in recent times, encompassing imaging techniques, image pre-processing or enhancement, parasite detection, feature computation, segmentation, and automated cell classification. An exhaustive amalgamation of preceding research endeavors, meticulously outlined for both thin and thick blood smear images.