Consider two random vectors A and B. Suppose they are equal in size. Let C and D be linear combinations of the elements of A and B respectively. The coefficients should be such that the correlations between C and D is maximum under the constraint that the variance of C and D are each equal to 1 and the additional constraint that C is uncorrelated with another linear combination of A and D uncorrelated with another linear combination of B.
- Canonical correlation analysis relates the two sets of variables by creating weighted linear composites for each of them in a set of predictive equations (functions) in the general configuration: dependent variables weighted linear composite = predictor or covariate variables weighted linear composite (Meyers et al., 2017).
- Use the General Linear Model module with a discriminant function analysis
Canonical analysis is a multavariate method used to find relationships between groups in a data set. You could find a brief explanation on how to perform canonical analysis in the following link:
https://onlinecourses.science.psu.edu/stat505/node/65/
- Badges
- Science topic
- Similar topics
- Medicine
- Physiology
- Biosignals
- Biomedical Signal Processing
- Statistical Analysis