CNNs have shown much better performance than fully-connected neural networks when dealing with images. In addition, CNNs are also capable of accomplishing good results with text and sound data.
CNNs have been explained in depth, as have how convolutions work and all the parameters that come along with them. Afterward, all this theory was put into practice with an exercise.
Data augmentation is a technique for overcoming a lack of data or a lack of variation in a dataset by applying simple transformations to the original data in order to generate new images. This technique has been explained and also put into practice with an exercise and an activity, where you were able to experiment with the knowledge you acquired.
Transfer learning is a technique used when there is a lack of data or the problem is so complex that it would take too long to train on a normal neural network. Also, this technique does not need much of an understanding of neural networks at all, as the model is already...