What are the Metrics used to Evaluate the performance of Regression Models in ML DM by Mahesh Huddar

[@RandhiGowtham]

very structured and informative, thanks bro

[@BijoyKarmakarBITSPilani]

Excellent explanation

[@MaheshHuddar]

Glad it was helpful!
Do like share and subscribe

[@mihawk_san]

It would be great if you also make another video showing what each value implies. For example, a value of 62 for MSE shows that the models is...
Thank you.

[@abdulhameedmalik4299]

Dear this is good lecture

[@MaheshHuddar]

Thank You
Do like share and subscribe

[@braincall2434]

Thank you 💓 I got pass exam

[@MaheshHuddar]

Welcome
Do like share and subscribe

[@MaheshHuddar]

In which university you are studying...?

[@KHANAZHARMEHMOODFAZALMEHMOODSt]

nice video

[@MaheshHuddar]

Thank You
Do like share and subscribe

[@alaataktokani8884]

thanks

[@parneetkaur-gs4bk]

Thankyou

[@MaheshHuddar]

Welcome
Do like share and subscribe

[@tamannakhan2579]

Why we just cannot use difference between actual value and predicted value?

[@prasad.kenkre]

I have the same doubt

[@just.devansh0201]

@@prasad.kenkre If we simply use the difference between the 'actual' and 'predicted' values, we might run into a situation where the 'negative' errors are balanced by the 'positive' error.
Say for example, we predict a price of 80 for something worth 75 and then we predict a price of 85 for something that is 90.
So, we have a '-5' error followed by a '+5' error, which when mean-ed, will result to a ZERO (0). That shows we have no error and our model is perfect. But, is it?
This is why, to avoid the zero-ing out of the sum, we use methods to remove the 'negative' sign from the error. (MAE, MSE, etc.)
I hope this clarifies.

[@prasad.kenkre]

@@just.devansh0201 Thanks but somehow I understood it partially

[@theseeker20210]

What is it that you're having trouble with?

[@binitapaul021]

Thnk u 🥹

[@MaheshHuddar]

Welcome
Do like share and subscribe