Algorithm

A set of rules or instructions given to an AI system to help it learn on its own.

Definition

A precise, step-by-step computational procedure—ranging from simple decision trees to complex neural-network training routines—that defines how an AI system learns patterns or makes decisions.

Real-World Example

In email spam detection, a Naïve Bayes algorithm calculates word-frequency probabilities to classify messages as spam or not, enabling engineers to trace misclassification to specific feature weightings when new spam tactics emerge.