How does reordering the features impact model training and its performance?

Per my understanding, it should not impact the model performance as weights get tuned according to feature value and not per the feature order.

Interested to know from others as well.


It depends on the type of dataset you have. For instance, if you are trying to classify different types of flowers, the order in which you train the model on the features is irrelevant.

However, that being said, if you are dealing with non-stationary data such as time series values i.e. stock market prediction, the order of the features relative to time, is very important.

  • $\begingroup$ Understood, but when I trained a model with a particular sequence, I'm getting comparatively good results. That's the reason I post this question. $\endgroup$ – vipin bansal Jun 6 '19 at 13:00
  • 1
    $\begingroup$ I see. What type of model/package are you using? Some models order the data based on the input. I.e. they can detect a time-series index in a pandas dataframe for example. $\endgroup$ – Cihan Dogan Jun 6 '19 at 13:03
  • $\begingroup$ quora.com/What-are-stationary-and-non-stationary-series $\endgroup$ – Cihan Dogan Jun 6 '19 at 13:04
  • $\begingroup$ Its not a time series data $\endgroup$ – vipin bansal Jun 6 '19 at 13:40

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.