Leveraging Deep Learning Techniques for Malaria Parasite Detection Using Mobile Application

Table 2

Model configuration summary including hyperparameters.

Parameter	Value/type
Epochs	50
Batch size	32
Optimizer	SGD with momentum 0.9
Learning rates	Min , max
Loss function	Categorical cross entropy
Input shape
Pooling	Max (convolutional layers), GlobalAverage (flatten layer)
Activation	ReLU (convolutional layers), Softmax (final dense layer)
Dropout rate	0.15
Trainable parameters	409,146