What is the good way working with 'Age' attribute? Don't touch it or should it be scaled? Below photo shows my results 'Before' and 'After' standardization.
$\begingroup$
$\endgroup$
4
-
2$\begingroup$ The answer depends on what you are using the attribute for. Input to neural network? Input to linear model? If linear, do you expect the model to be interpretable? $\endgroup$– noeCommented Jan 22, 2023 at 17:27
-
$\begingroup$ My attribute is from dataset, this dataset is used for learning a machine learning model (s), like simple ones. And I'm not sure if the model is going to be interpretable. $\endgroup$– hyper-cookieCommented Jan 22, 2023 at 17:31
-
2$\begingroup$ Different scales of the dependent variables affect the interpretability of feature importance in linear models, that's why you normally transform them into the same value scale. Therefore, if you are going to use it in a logistic regression where you need interpretability. $\endgroup$– noeCommented Jan 22, 2023 at 17:47
-
$\begingroup$ does standardization in your case (graphs) change the meaning of data ? what is the formula you apply for it? $\endgroup$– Subhash C. DavarCommented Feb 19, 2023 at 3:25
Add a comment
|