Question: How to assess information content / mutual information between 2 sets of variables ?
gravatar for Khader Shameer
5.0 years ago by
Manhattan, NY
Khader Shameer18k wrote:

Am testing data from a clinical device; data was extracted using two complementary methods 

  •     Method 1: Existing method, 17 variables 
  •     Method 2: New method, 13 variables 

I need to quantitatively and visually explain that the **information content / mutual information** of both methods are the same. 

I have correlation data but I am looking for an intuitive way to explain that irrespective of differences in number of variables between the data extraction methods and direction of correlation the information captured by methods are comparable. 

Please let me know your thoughts, thanks in advance !

Cross-posted from here

data-science • 1.7k views
ADD COMMENTlink modified 5.0 years ago by Ahill1.6k • written 5.0 years ago by Khader Shameer18k
gravatar for Ahill
5.0 years ago by
United States
Ahill1.6k wrote:

If you showed by linear regression that each of the 17 existing method variables was approximately equal to some linear combination of the 13 new variables, that would be fairly convincing that you've captured the same information in both methods.

ADD COMMENTlink written 5.0 years ago by Ahill1.6k
Please log in to add an answer.


Use of this site constitutes acceptance of our User Agreement and Privacy Policy.
Powered by Biostar version 2.3.0
Traffic: 811 users visited in the last hour