Deep Learning and Convolutional Neural Networks; Layers of a CNN: convolution, pooling, fully-connected layer; Convolution: padding and stride; Loss functions; Forward and back propagation; Parameters and hyperparameters; The training process: activation functions, pre-processing, weight initialization, optimization; Optimization: SGD, Momentum, Adam; Regularization: dropout, data augmentation, batch normalization, weight decay, early stop; CNN architectures; Applications and advanced topics.