Generalization in QML from few training data

what is the fairest way to check the interesting result of this code with a classical convolution neural network. Because comparing a quantum and classical algorithm in ML is the hardest task as far as i have read in this paper: Key questions for the quantum machine learner to ask themselves - IOPscience.

1 Like

Hey @sassan_moradi, welcome back!

When comparing classical and quantum NNs, answering what is “apples to apples” is tricky. But, I’d say that so long as you’re processing the data the same way (i.e. the input data to both is the exact same), validating/training both models similarly, and are able to have reproducible results, you’re off to a good start!

many thanks for your useful comment!

1 Like