Principal Components and Independent Components

Extreme events: blending PCA and ICA [25]

Go to: Summary | Previous | Next
Bullet points include: Fat tails involve deviation from Normality. Hence at least some of the higher cumulants (moments), aka semi-invariants, of the distribution, e.g. skew and (excess) kurtosis, must deviate from zero (Normality). Use of these (and possibly other higher cumulants) is most common way of analysing and coping with fat tails, but it is not necessarily the best approach

Bullet points include: Fat tails involve deviation from Normality. Hence at least some of the higher cumulants (moments), aka semi-invariants, of the distribution, e.g. skew and (excess) kurtosis, must deviate from zero (Normality). Use of these (and possibly other higher cumulants) is most common way of analysing and coping with fat tails, but it is not necessarily the best approach

NAVIGATION LINKS
Contents | Prev | Next | Library

Desktop view | Switch to Mobile