Supplementary

The supplementary notebooks are written to be approachable, but they assume some comfort with the usual mathematical and programming background for an artificial intelligence course.

See the full prerequisites page for a compact checklist and recommended review material.

These notebooks are gateway tutorials for selected artificial intelligence topics: short enough to browse, but practical enough to include code, examples, visualizations, and pointers for going deeper.

Mathematical and Classical Foundations

• Gradient Descent in Machine Learning: Gradient descent, practical challenges, and common improvements.
• Convex Optimization: Convex optimization fundamentals with a practical image in-painting example using CVXPY.
• Decision Trees: Decision trees, pruning, and indexing ideas in classical machine learning.
• Clustering in Machine Learning: Clustering variants and their behavior in unsupervised learning workflows.
• Dimensionality Reduction: PCA, t-SNE, UMAP, and LDA on MNIST for visualizing lower-dimensional structure.
• Machine Learning Methods: Online learning, semi-supervised learning, few-shot learning, and meta-learning.

Logic, Sequence Modeling, and Language

• Prolog and Logic Programming: Prolog basics and how to use logic programming ideas from Python.
• HMM Speech Recognition: Hidden Markov Models for speech recognition, with useful automatic speech recognition tools.
• Recurrent Neural Networks: Sequence modeling with recurrent neural networks.
• Natural Language Processing: NLP challenges and approaches for working with text.

Deep Learning, Vision, and Generative Models

• Introduction to Keras: A practical introduction to Keras for deep learning workflows.
• Convolutional Neural Networks: CNNs, convolution, pooling, image classification, and TensorFlow training.
• Transformer: Attention, positional encoding, encoder-decoder structure, and transformer sequence modeling.
• 3D Computer Vision: Point clouds, 3D vision neural networks, and adversarial attacks.
• Generative Adversarial Networks: GAN fundamentals, applications, and an anime face generation example.
• Generative Adversarial Networks with WGANs: GANs, conditional GANs, WGAN variants, and PyTorch implementations.
• Robust and Trustworthy Machine Learning: Evasion attacks, poisoning attacks, and defenses for robust ML systems.

Systems, Applications, and Reinforcement Learning

• Introduction to Big Data: Big data concepts, distributed data processing, and Apache Spark in ML workflows.
• MLOps: A simple model deployment walkthrough and the engineering ideas around MLOps.
• Deep Reinforcement Learning: Early deep-learning approaches to reinforcement learning, including a DQN example.
• AI Applications in Global Warming: How AI can support climate analysis, including CO2 emission prediction.

If you want to go deeper than the notebooks, we keep a curated list of major public courses and references across computer vision, unsupervised learning, reinforcement learning, deep learning, and related topics.

See the full further reading page for the course list.