I'm trying to figure out how I can use JMP to solve the following problem. I have a data set with a bunch of values in each row. What theoretical technique/how do I use JMP to determine if there are any relationships between 1 or more values. (for example, if I have A in the record, you will most likely have B, C, and D as well.
The rows would be encoded like this:
A B C D E A B C Q R S T A D D A E M R Z R S T L N E A B A D E A B D E
Here is a first try that examines correlations among values and trys a little clustering. One can treet the problem like text mining. Once you have your "document by term" matrix, there are many ways to examine relationships among values.