[Task] Create "Decision" Tree Page to Guide Users to the right library/model/examples

Description

This is similar to Scikit Learns "Choosing the right Estimator":
https://scikit-learn.org/stable/tutorial/machine_learning_map/index.html

We would like to have a similar document, explaining different decisions for modelling/choosing the correct library.
I think the document should be added to the main Documentation:
https://nvidia-merlin.github.io/Merlin/main/README.html

Some examples questions could be (but not limited):

• Do we have a binary target or multi-class classification (next item prediction)
• Do we have a timestamp column and a user column -> session based
• Do we have >1M items for NextItemPrediction -> Retrieval Model / Negative Sampling required
• Do we have large embeddings -> SOK, Distributed Embeddings
• etc