Options

How to find the training accuracy from my process here?, normally the output just the test accuracy

Christian18Christian18 Member Posts: 3 Newbie
in Help

Answers

  • Options
    rfuentealbarfuentealba Moderator, RapidMiner Certified Analyst, Member, University Professor Posts: 568 Unicorn
    Hello @Christian18,

    Your training accuracy will be normally high because it is your training data, and your model will fit to that. Therefore it's the testing accuracy (the one you already got) the most important value you should consider.

    BTW you could basically replace the testing data to the training data and see how it performs, but there might not be a lot of interest in doing that (unless you do that and get totally wrong results; if such a thing happens, you should try applying normal statistics to your training data to see how imbalanced it is and whether it has null values or too many unique ones, correlations are too low... you name it).

    Hope this helps.
Sign In or Register to comment.