Supervised, unsupervised, and deep learning, generative models, reinforcement learning, and adversarial search — the algorithmic core behind modern AI.
"Regression" is a strange name for a family of techniques used to predict continuous values, and the strangeness has a specific origin: Francis Galton, studying the heights of parents and children in the 1880s, found…
Supervised learning starts from labelled examples — every input comes with a known correct answer to learn from. Unsupervised learning starts from data with no labels at all, and asks a different kind of question: what…
Every architecture below solves the same underlying problem — learning a function from examples, by adjusting internal numbers until the function's output gets closer to what's wanted — but each one earns its place in…
A classifier answers "what is this?" A generative model answers a different question entirely: "produce me a new, plausible example of this." Three architectures have dominated that second question over the last decade…
Supervised learning is taught a correct answer for every example it sees. Reinforcement learning is taught nothing of the kind — an agent takes actions in an environment, receives a reward signal that only ever says…
Every search algorithm covered on Divide and Conquer and its neighbouring pages assumes a single agent moving through a static problem. Adversarial search drops that assumption: an opponent is also choosing moves, and…