Generalized Canonical Correlation Analysis: A Subspace Intersection Approach

Generalized Canonical Correlation Analysis (GCCA) is an important tool that finds numerous applications in data mining, machine learning, and artificial intelligence. It aims at finding `common' random variables are strongly correlated across multiple feature representations (views) of the same set entities. CCA to a lesser extent GCCA have been studied from statistical algorithmic points view, but not as much standpoint linear algebra. This paper offers fresh algebraic perspective based on (bi-)linear generative model naturally captures its essence. shown algebra point tantamount subspace intersection; conditions under which common different views identifiable provided. A novel algorithm proposed intersection, scales up handle large tasks. Synthetic well real experiments provided showcase effectiveness approach.