ABSTRACT
Canonical Correlation Analysis (CCA) is used as a dimensionality reduction technique for multi-view setting, that is, when there are two or more views of the same data. Whenever we have two or more views of the same data points, and we are interested in the information in all the views, we are interested in the redundant information provided by all the views. Redundant information can be used to reduce the noise in either one or both the views and obtain a lower dimensional representation of the data that consists of redundant information. This chapter discusses how to generate low dimensional representations of the points in each view such that it retains the redundant information between the different views by singular value decomposition. Alongside, this chapter works through the math and the underlying theory of CCA and discusses its advantages, limitations, and use cases.
