I have a philosophical question here. What are the benefits, or why it is necessary to use benchmark/baseline models in any machine-learning effort?

For example, if I use a simple benchmark model against a complex deep learning model, the result might tell me that for this dataset and in this application, the use of the complex model is not worth it, or justified.

But what else we can learn from benchmarking?



Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy