<![CDATA[Malaria is a type of disease caused by protozoan cells called Plasmodium, in its development Plasmodium cells will enter and destroy red blood cells. Detection of malaria is difficult because it takes a long time. Malaria parasite can be identified using a microscope to see whether or not Plasmodium is present in red blood cells. However, this method is very dependent on the quality of the microscope. A computer vision model was developed using a CNN (convolutional neural network). The model was developed by comparing the architectures of the ResNet-101, AlexNet, and VGG-19 models by providing two learning rate scenarios, namely the minimum learning rate and the maximum learning rate with slices. The training process for each model uses 10 epochs. Fast.ai tools / libraries are used to form existing models. The results of a study entitled Plasmodium Image Classification To Detect Malaria Disease by using the CNN Algorithm found that the architecture of the ResNet-101 model has much better accuracy than the AlexNet and VGG-19 models, both by using a minimum learning rate and a maximum learning rate, the process of training in the ResNet-101 architecture model using the maximum learning rate with slice has the best accuracy of 0.97586% and precision of 0.98249% compared to the AlexNet and VGG-19 architectures.]]>
Copyrights © 2022
