3. Machine learning

Machine Learning (ML) allows a system to discover patterns from historical examples instead of requiring every decision rule to be manually written.

Historical data + Correct answers → Training → Model
New data + Model → Prediction

Suppose you have historical support tickets:

Instead of maintaining hundreds of if statements, an ML algorithm analyzes these examples and produces a classification model.

Classification means predicting a category.

Other common ML tasks include:

• Regression: Predicting a number, such as house price.
• Classification: Predicting a category, such as fraud/not fraud.
• Clustering: Grouping similar records without predefined categories.
• Forecasting: Predicting future values over time.
• Recommendation: Ranking products or content for a user.
• Anomaly detection: Finding unusual behavior.