This example uses the Santander Customer Transaction Prediction data, from
https://www.kaggle.com/c/santander-customer-transaction-prediction/data.  This
is a 200,000 record, 200 column dataset.  The dataset is stripped of the string
ID_code, scaled and split into two parts for training (180,000 records) and
testing (10% or 20,000 records).

The data is stored as prepositioned Financial Tabular Data. Said data is split
into three parts. Each data file is placed on an independent Access Point,
Representing three private bank infrastructures. The column named "target"
is the "y" label, and the rest of the columns get stored under the "x" data.

These datasets are used by organization 3 to perform a distributed and
private XGBoost model training. The resultant model resides on org 3's
Access Point and can be used by others with permission for a remote inference.
The model can also be download by org 3 for local inference.

Both federated and SMPC inference are showcased.
