Back to Glossary
Machine Learning (ML)
Fundamentals
Algorithms that learn patterns from data instead of being explicitly programmed.
Machine Learning optimizes model parameters so that predictions generalize from examples.
- Main types: Supervised, unsupervised, and reinforcement learning.
- Workflow: Data preparation → training → validation → testing/monitoring.
- Risks: Overfitting, data leakage, distribution shifts (data/concept drift).
- Best practices: Proper data splits, baselines, reproducible pipelines, and task-appropriate metrics.