-4

I recently conducted a small hackathon , not on a platform like kaggle , but only provided the participants with the training data, and the test data without the true labels.

Is there a way in which I can evaluate their submissions?

user42868
  • 1
  • 2
    how about accuracy score? see https://towardsdatascience.com/metrics-to-evaluate-your-machine-learning-algorithm-f10ba6e38234 – Nikos M. Jul 05 '20 at 08:53

1 Answers1

0

You split you train data into train, val and test data. You don't have to use this test data anywhere in the training. It will behave similarly to your actual test data. Run evaluations on this dataset.

Patel
  • 200
  • 2
  • 9
  • this answer has nothing to do with the question which asks how to evaluate already made results, not training models!! – Nikos M. Jul 05 '20 at 09:41