... signs.150 WHAT IS INDEPENDENT COMPONENT ANALYSIS?s1(t)s2(t),ands3(t)are, at each time instantt, statistically independent.This is not an unrealistic assumption in many cases, and it ... which is typically called whitening or sphering,and often implemented by principal component analysis. In this section, we show thatthis is not possible, and discuss the relation between ICA and ... GAUSSIAN VARIABLES ARE FORBIDDENWhitening also helps us understand why gaussian variables are forbidden in ICA.Assume that the joint distribution of two ICs,s1ands2, is gaussian. This means...