Options

# "multivariate correlation - groupwise correlation"

Hi everyone,

I want to compute the correlation coefficience between 2 or more groups of attributes. Mathematically, I have 2 groups: X and Y. X has 2 attributes (A, , Y has 3 attributes (C, D, F). How can I compute R(X,Y)?

Tagged:

0

## Answers

1,635UnicornI do not believe there is a standard definition of correlation between asymmetrical groups of attributes, since there are several ways that it could be done based on the underlying correlations of the individual attributes. Is there a reason you need to aggregate these at the group level rather than just looking at the correlation matrix for (A,B,C,D,F) which is easily produced in RapidMiner with the Correlation Matrix operator?

Lindon Ventures

Data Science Consulting from Certified RapidMiner Experts