Auto Encoder Operator specifications

nakashi

Nice to see you here.

I have a question regarding the Auto Encoder Operator specifications.

We are currently checking to see if there are any differences in the models output from the autoencoder operator between RapidMiner versions 9.10 and 10.1.

After verification, we have confirmed that there are differences in the weights and biases of the hidden layers.

So my question is, is there any way to prevent such differences in the hidden layers between versions? (e.g., replace the operator part of the autoencoder with Python code).

If you need more detailed data or output results, please let me know.

Thanks for the confirmation.