Topics covered in this 2-hour session on Identifying Data Relationship
- Data Qualification
- One 2 one
- One 2 many
- Many 2 many
Part1: Data Qualification
Before exploring the relationships between the data, it's important to understand them and look at their structure, particularly for Y responses. Distribution analysis is an essential first tool
Part2: One 2 one
Thanks to its interactivity, the distribution tool can be used to identify certain relationships that can be validated by the Fit Y by X tool, which adapts to all situations (continuous/categorical X, continuous/categorical Y, etc.).
Part3: One 2 many
When the number of variables increases, and so does the complexity of the relationships, the one 2 one relationship is no longer sufficient. Hence the need to look for influential variables using tools such as the partition tree or predictor screening.
Part4: Many 2 many
Get an idea of the overall relationships between the data, which is what correlation matrices allow you to do.
