search

juliasilge / juliasilge.com

My blog, built with blogdown and Hugo :link:
https://juliasilge.com/
40 stars 27 forks source link

Modeling #TidyTuesday NCAA women's basketball tournament seeds | Julia Silge #69

Open utterances-bot opened 2 years ago

utterances-bot commented 2 years ago

Modeling #TidyTuesday NCAA women's basketball tournament seeds | Julia Silge

A data science blog

https://juliasilge.com/blog/ncaa-tuning/

conlelevn commented 2 years ago

Hi Julia, In this screencast I have learned a new approach for non-linear relationship bwt prediction and target variable, usually when I see that kind of relationship I will try to log the prediction or target or both of it in order to seeking a linear relationship. In your view, which approach is optimal for this kind of data? SPLINE or transformation?

juliasilge commented 2 years ago

@conlelevn I could see trying out a log transformation, like log(seed) or similar, since this is not too "wiggly" of a curve. With predictors that change in more complex way, like this example, you would want to use splines instead.