When To Use Classification or Regression?

[@OscarAlsing]

🎉Don't forget to subscribe if you want to engage more in similar content! 🎉
✨Support me on Patreon: patreon.com/oscaralsing
✨Join us on Slack: oscaralsingcommunity.slack.com/join/shared_invite/enQtMzA1NTIyNTgxMTI0LWQ2YmFiZTM5MWE5NDcxYmY0OGU5ZWM3Y2VmMWI1N2RjNjQzYTdlMzNkNmNlYzI2YzU4OTUyYmM0OGMwZDlkYjE

[@hammadshaikhha]

Can you make a video about how steemit works?

[@OscarAlsing]

hammad shaikh I'll have a look!

[@hammadshaikhha]

Great video, I think this is also a tricky question because a regression can also be used for classification. If Y is 0 or 1, then the predicted values from the regression Y = b0 + b1X give us an estimate of P(Y = 1|X). The classifier can be Y = 1 if predicted value of Y > 0.5. But this classifier has several weaknesses and wont perform as well as our standard SVM.

[@OscarAlsing]

hammad shaikh wonderful addition! ❤️

[@limgg7917]

@hammadshikhha, what you have mentioned is logistic regression. You may not want to confuse this with linear regression, as logistic regression is a special case, purely for classification or probability. Btw, it's logit function, not Y. Your equation is not right too.

[@rag_llm]

Thanks for clarifying the difference between the two techniques Oscar!

[@Somuchgam]

"Pepsi or cola" *Pam looks at both* "These are the same picture"

[@stefancronje6202]

Great explanation, would love to learn more about AI programming and how it works. Im comfortable with high level maths and programming in MS Office (macros etc)

[@kh7227kk]

good explanation ! thanks

[@rhsa6292]

killer video dude . keep it up

[@bishalpaul1826]

But logistic regression also gives binary output label

[@sarahaghighi1382]

Thank you Oscar

[@clevo4040]

"Length of a 14-year old boy". That just sounds wrong. :)

[@OscarAlsing]

Haha never though about that

[@Anamaziz1233]

informative

[@gamer7234]

12 bathrooms!?