Deep Learning & Neural Networks

Deep learning is the foundation beneath modern language and vision systems. These courses build neural networks from the ground up, covering backpropagation, activation functions, loss design, regularisation and the practical craft of training: learning rate schedules, batch size, initialisation and diagnosing a model that will not converge. Architectures range from convolutional networks for images to recurrent networks for sequences and the transformer that displaced them. Expect hands-on work in PyTorch, TensorFlow and Keras, with GPU training and transfer learning covered. Take this before specialising in NLP, computer vision or model fine-tuning.