When new to CNN and deep learning we often find it difficult to find correct resources to get stated. This code will hel you get started in the easiest manner and will help you gain poerspective of how CNN works and different elements of CNN
To get started please download the data from:
Step-1
To get started we need to have a google colab account or just sign in to google colab
Step-2
Place the downloaded data to a place in gDrive mine was stored as cell_images.zip in Malaria-Detection folder in my gDrive
Step-3
We need to unZip and bring the data in colab environment
Step-4
Read in the data and store it in array format. Here, you can choose whatever shape you like, I have restricted it to 64x64. If we increase the image size the total file size will increase drastically and we might run out of resources to while running our CNN Model
Step-5
After conversion of the dataset and getting final array, we save it for further use.
Step-6
Define a CNN model based on your learning or exploration. Add as Conv2d layers, Dropouts and MaxPooling layers, you can also try padding and stride. Just check the hyperparameters for Conv2D
Step-7
Basic implementaion of model has been done in the code, you can play around with # of Epochs or batch size or the optimizer altogether.
Please go through the attached code in detail to get the idea of CNN, Malaria Detection using Google Colab