Source: Machine Learning Mastery
Published: March 2016
Supervised and Unsupervised Machine Learning
Circulated: January 3, 2020
Supervised learning is like learning with a teacher. From historical data, we know correct answers that have happened. There two types:
A classification problem is when the algorithm predicts a category, like an email is spam or not spam.
A regression problem is when the algorithm predicts a value, like how many minutes it will take to drive home from work.
Unsupervised learning is when there are no correct answers and there is no teacher. Algorithms are left to discover interesting structure in the data you provide. There are two types:
A clustering problem is where you want to group data, like separating news articles into a specified number of topics.
An association problem is where you want to find rules that are generally true, like customers who buy bagels also tend to buy cream cheese.