Generative vs. discriminative models in ML: (a popular ML interview question)

@_avichawla Nicely visualised, thanks for sharing Avi!

@_avichawla Interesting framing. One creates, the other critiques. Are they two sides of the same coin?

@_avichawla Generative models are like the artists of ML, while discriminative models are the critics. Both have their place, but one’s definitely more fun at parties. #PublicAI

@_avichawla Generative models are like the artists of ML, while discriminative models are the critics. Both have their moments, but one’s definitely more fun at parties!

Generative models learn the joint distribution so they can create data that looks real, like GPT writing a paragraph. Discriminative models learn the boundary between classes, like a spam filter deciding if an email is spam or not. One imagines the world, the other draws lines through it.

@_avichawla Knowing the difference can really help in picking the right approach for a problem

@_avichawla Just a stupid question, Which tool do you use for making this elegant graphs? Thanks a lot in advance