A New decomposition approach for binary product series via high dimensional modeling.

This paper aims at bringing a new point of view to binary products series via Tridiag- onal Kernel Enhanced Multivariance Products Representation (TKEMPR). For this purpose a new version of TKEMPR has been developed such that each additive term of binary products series is represented using bivariate EMPR by starting with appropriate weight functions, and, left and right support functions. Hence we call it "Overlapped Square Blocks Tridiagonal Enhanced Multivariance Products Representation (OSBTKEMPR)". The basic goal of OSBTKEMPR is not to get an efficient approximation scheme since its own structure, as a matter of fact, is based on truncated terms. The main purpose is to reveal the connections between the binary product series and OSBTKEMPR, and then, achieve certain practicalities. [ABSTRACT FROM AUTHOR]