With the idea of principal component analysis, CCA reflects the linear correlation between two sets of variables with a few basis vectors.

• 它借助主成分的思想，用少数几对综合变量来反映两组变量间的线性相关性质。