0
$\begingroup$

I've been watching a few training videos from Orange here and attempted to reproduce the process. They used iris dataset for classification task. When I compared my confusion matrix to theirs, I didn't get the same results.

Is this a problem with Orange software or with sklearn (I know they somehow leverage sklearn)? When you run a code again 6 years later, you get different results, even though the dataset is the same...

$\endgroup$

1 Answer 1

1
$\begingroup$

In general, algorithms has a lot of maths behind. Maybe the difference between Orange software and sklearn is due to small differences in these maths. Of course, only small differences should appear.

Moreover, many algorithms (like random forest) are created with some type of randomness; for example, random forest select randomly samples and features. So here is another possible factor that creates differences.

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.

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