Investigate Keras Model Compression before SQL generation

[antoinecarme]

This is probably a way to avoid some unnecessarily large models and generate an (almost) equivalent SQL code.

[antoinecarme]

There are at least 4 methods for compressing keras models in

https://arxiv.org/pdf/1710.09282.pdf

[antoinecarme]

[antoinecarme]

Would be nice to test this :

Model Compression Based on Geoffery Hinton's Logit Regression Method in Keras applied to MNIST 16x compression over 0.95 percent accuracy

https://github.com/Irtza/Keras_model_compression

[antoinecarme]

Factorization method :

low-rank approximation 1. with SVD (matrix) 2. with Tucker (tensor)

https://github.com/DwangoMediaVillage/keras_compressor