Constrained clustering

Constrained clustering is a way of organizing data into categories. Basically, it is like putting items into different piles. A constraint is like a rule that helps you decide which items should go into which piles. For example, if you were sorting different kinds of clothing (like shirts, pants, and socks) you might make a rule that all white clothing goes in one pile and all colored clothing goes in another. That rule is a constraint, because it tells you which pile each item should go into.