Общие понятия

WARNING

СТАТЬЯ В АКТИВНОЙ РАЗРАБОТКЕ

О чём писать

Объекты и признаки, классификация задач и подходов (с учителем и тд, классификация и тд), примеры задач

---

Machine learning problems • Supervised learning • Unsupervised learning • Semi-supervised learning • Reinforcement learning • Active learning • Online learning • Structured prediction • Model selection and validation

Supervised learning A set of examples with answers is given. A rule for giving answers for all possible examples is required: • classification; • regression; • learning to rank; • forecasting.

Unsupervised learning A set of examples without answers is given. A rule for finding answers or some regularity is required: • clustering; • association rules learning; • recommender systems*; • dimension reduction**.

How are the objects described? f j ∶ X → D j , j = 1, ... , n are features or attributes. Feature types: • binary: D j = 0, 1 ; • categorical: D j is finite; • ordinal: D j is finite and ordered; • numerical: D j = R.