Applying deep learning to Related Pins
Pinterest Engineering

I understand that the the size of the input layer is the size of the vocabulary, number of pins. So is the output, right? My question is, how do you deal with new pins added to your system every moment? Do you need to re-train the model continuously, by adapting the one-hot-encoding to include all the newly added pins?

Like what you read? Give Tarek Amr a round of applause.

From a quick cheer to a standing ovation, clap to show how much you enjoyed this story.