search

pierz95 / SpotifyAttributes

2 stars 0 forks source link

Motivation for discretising personality traits into three groups #1

Open NikolayBlagoev opened 4 months ago

NikolayBlagoev commented 4 months ago

What was the motivation for discretising personality traits into three groups with the said decision boundaries? This is not a typical psychological distinction. Why opt for that instead of for continuous variable and train a regressor?

pajola commented 4 months ago

Thanks for reaching out. We agree with your comment. In our project, we only decided to work mainly on classification tasks, and therefore, we discretized all the target continuous variables. But the choice of using a regression makes sense and it would be a better choice.