A complete, self-paced path through classical machine learning: statistics and problem framing, data wrangling and feature engineering, regression, evaluation, classification with ensembles, unsupervised learning, model-quality pitfalls, and a taste of reinforcement learning — finishing with a hands-on capstone project.
A hands-on, self-paced path through neural networks: start at a single neuron and backpropagation, work through optimization and regularization, build CNNs, RNNs, and attention/transformer architectures from scratch, and pick up the practical skills to train, debug, and avoid common pitfalls — finishing with a deep learning capstone project.