Blending Independent Components and
Principal Components Analysis
Abstract and contents
[this page | pdf | references | back links]
Abstract
In these pages we describe the similarities and differences
between independent components analysis and principal components analysis,
particularly as applied to portfolio risk model construction. We describe how
the best elements of each approach can be blended together to enhance the
ability of portfolio risk models to handle fat-tailed behaviour.
Contents
1. Introduction
and Conclusions
2. Independent
Components Analysis (ICA)
3. Principal
components analysis (PCA)
4. Blending
together PCA and ICA in the construction of risk models
Tools
References
NAVIGATION LINKS
Contents | Next