Run some Covid-19 ICU predictions via ML vs. IntegratedML

Keywords:  IRIS, IntegratedML, Machine Learning, Covid-19, Kaggle 

Purpose

Recently I noticed a Kaggle dataset  for the prediction of whether a Covid-19 patient will be admitted to ICU.  It is a spreadsheet of 1925 encounter records of 231 columns of vital signs and observations, with the last column of "ICU" being 1 for Yes or 0 for No. The task is to predict whether a patient will be admitted to ICU based on known data.

This dataset seems to be a good example of what's called "traditional ML" task. The data seem to have the right quantity and relatively right quality. It might have a better chance of being applied directly on the IntegratedML demo kit, so what could be the simplest approach for a quick test based on normal ML pipelines vs. possible IntegratedML approach?

Scope

We will briefly run through some normal ML steps such as :

• Data EDA 

• Feature selection

• Model selection

• Model parameter tuning by grid search

Vs. 

• Integrated ML approaches via SQL.

It's run on an AWS Ubuntu 16.04 server with Docker-compose etc.