search

dreamquark-ai / tabnet

PyTorch implementation of TabNet paper : https://arxiv.org/pdf/1908.07442.pdf
https://dreamquark-ai.github.io/tabnet/
MIT License
2.65k stars 488 forks source link

Generate Embeddings for Tabular Data #521

Closed Gaurav7888 closed 11 months ago

Gaurav7888 commented 1 year ago

Feature request

Generate Embeddings For the Tabular data

What is the expected behavior? We can get embeddings of tabular data

What is motivation or use case for adding/changing the behavior? It can be used in several other use cases.

How should this be implemented in your opinion? If tabnet is encoder-decoder type model then just throw the vector embeddings generated after encoder layer

Are you willing to work on this yourself? No

eduardocarvp commented 1 year ago

The encoder layer returns steps_out which is then input to the decoder:

steps_out, _ = self.encoder(masked_x, prior=prior)
res = self.decoder(steps_out)

Is that what you mean by the vector of embedding generated?