EHR Transformations & Feature Engineering Overview

Transformations and Feature Engineering Overview

This lesson is divided into 3 parts:

1. EHR Dataset Levels

In this part, there are three levels - line, encounter, and longitudinal. By the end of this section, you will be able to identify the level of your dataset as well as conduct tests and transform your data.

2. Dataset Splitting Without Data Leakage

In this part, you will learn about dataset splitting without Data leakage, which can be a major issue in EHR datasets. By the end of part two, you will be able to implement some basic tests to help prevent issues when splitting data.

3. Feature Engineering with Tensorflow

Finally, we will cover Feature Engineering with Tensorflow. In this part, we will cover ETL (Extract, Transform, Load) using TensorFlow. This will allow you to scalably process and transform your data for modeling. You will also be able to transform datasets using the TF Feature Column API for both numerical and categorical features. The Feature Column API can be extremely useful for transforming datasets at scale and building some unique feature types.

Let's get started!