Should I need to convert ordered numerical data into categorical for encoding or normalize it?

Question

I'm working on a dataset that has both numerical and categorical columns. One of the numerical columns is fare rates ($) which has just 4 distinct values (200, 400, 600 and 800). I have done feature scaling on other numerical features but I'm stuck here to decide whether should I need to apply normalization here or make it categorical to encode this feature. I want to use Neural Networks, if I treat it as a numerical feature, the weights for this feature will affect the output. If anyone has any leads then please help me in this regard. Thanks

I'm trying to find a perfect solution.

score 0 · Answer 1 · answered Dec 06 '22 at 18:19

0

Use it as numerical feature, it should give you a better result. I

answered Dec 06 '22 at 18:19

jacl613 max31

306
3

Can I get to know the reason why should I use it as a numerical feature or why not use it as a categorical feature? – Muhammad Kaleem Ullah Dec 06 '22 at 19:03

Should I need to convert ordered numerical data into categorical for encoding or normalize it?

1 Answers1