Add redshift information to model

[tallamjr]

Comparing to results by Alves et al. Z information have a large importance weighting to the classification. Therefore, it seems promising to add redshift information to the existing T2 model.

This seems to be possible with https://datascience.stackexchange.com/questions/22632/adding-hand-crafted-features-to-a-convolutional-neural-network-cnn-in-tensorfl

[tallamjr]

Look at how best to combine inputs, maybe look at Siamese networks

[tallamjr]

Refs:

• https://stackoverflow.com/questions/48827687/add-hand-crafted-features-to-keras-sequential-model
• https://stackoverflow.com/questions/58153888/how-to-set-the-input-of-a-keras-subclass-model-in-tensorflow
• https://stackoverflow.com/questions/59743161/tensorflow-model-subclassing-mutli-input
• https://stackoverflow.com/questions/43196636/how-to-concatenate-two-layers-in-keras/43196972
• https://stackoverflow.com/a/20509316/4521950