Machine learning that uses hand-made features and older algorithms to learn patterns.
Classical Machine Learning is like a coach with a clipboard. It checks your stats before picking the team.
It works well on neat tables of data. You meet it in fraud checks and product picks. It also helps people study business data.
Supervised Learning
Classical Machine Learning often predicts from labeled data.
Unsupervised Learning
It can also find structure in data without labels.
Feature-engineering
Hand-made features often decide how good the old model gets.
Deep Learning
Deep Learning later reduced the need for hand-made features.