Hi, thanks for the video. Could you elaborate more on how to actually interpret the polynomial coefficients, other than the fact that you can't interpret them linearly, as you state at the end of this video (maybe with an example?) Thanks
So basically when you need to add a non-linearity to fit your model better, you can't interpret it in any meaningful way that our brain would intuitively understand. With the exception of logistic regression, probably.
Not at all... interpreting a curve is harder than a line, but anyone who throws a ball in the air to another person knows how to interpret the "arc" of the ball... some places it is going up, other places it is coming down. Very intuitive, no?
@@michaelkevane8210 Ah, thank you, I think it gets a bit clearer now. So in case of non-linear dependency, the interpretation just takes a different form than what we are used to in the linear case. So although we're not able to make an intuitive sense of it numerically, we can try to get the idea visually.
@@LittleLightCZ Exactly.... the numerical interpretation of how change happens is not so simple as just delta y over delta x = number... now "it depends"!
