Skip to content

Comment on You don't need ML/AI, you need SQL

Comments

Counting items by value is a maximum likelihood estimation method too. It's still ML if you do a count, group by, max or threshold - just a less sophisticated way of doing things. The Naive Bayes algorithm is implemented by counting, at its base.

So any reduction is ML?

Only if it fits to an existing statistical model.

How about counting the elements of a linked list.

AboutSource Built by g1lg1l

Hackerly is an independent reader for Hacker News, built on the public HN API. Not affiliated with Y Combinator.