0

I am using a simple CNN model, and I want to run a test, input is an image of Bird, the output is either "0" or "1", where "1" means the output is correct, is a bird.

Do I need to set my training data and validation image data? If yes, can I weigh it as 90% training data and 10% validation data?

Thank you.

aan
  • 75
  • 8
  • 1
    This is not dependent on CNN's, but rather what you wish to do with the model. If you want a model that interpolates the data you have already, then having only training data is fine. But if you want to make predictions of future (possibly independent) data, you will want to validate your model. – DifferentialPleiometry Oct 31 '21 at 22:40
  • @Galen. Thanks. Is there a difference between "validation data" and "test data"? – aan Oct 31 '21 at 22:45
  • @Galen Thanks. Is there a difference between "validation data" and "test data"? – aan Nov 01 '21 at 19:50
  • https://stats.stackexchange.com/questions/19048/what-is-the-difference-between-test-set-and-validation-set – DifferentialPleiometry Nov 01 '21 at 19:52
  • 1
    @galen, Thank you – aan Nov 06 '21 at 16:20

0 Answers0