I am working on stock price prediction project, I am using the support vector regression (SVR) model for it.

As I am splitting my data into train and test, I am getting high accuracy while predicting test data after fitting the model.

But now when I am trying to use another data which I separate out from the used dataset before start doing anything, it gives me very bad results. Can anyone tell me what's happening?

enter image description here

Looking forward to your response.

  • $\begingroup$ Rather than providing a screenshot of the code, it may be a good idea to explain how you obtained that separate dataset. Maybe you should also describe more in detail your datasets (e.g. how many entries they have, how many features, etc.). Also, note that in the source code that you're showing us, apparently, you're calculating the r2_score rather than the accuracy. You should clarify this (because in your post you say "accuracy"). $\endgroup$
    – nbro
    Commented May 25, 2020 at 23:50
  • 1
    $\begingroup$ The difference is very large. Have you used any type of scaling? Maybe you forgot to scale the newest data set? $\endgroup$
    – GKozinski
    Commented May 26, 2020 at 5:14


You must log in to answer this question.