Chapter Meeting: Independent Component Analysis

Things to do
0 Love It
Save it
Chapter Meeting: Independent Component Analysis
More Less
GARP says
Dallas, Texas Chapter Meeting

**You must register with GARP to attend**
Register here:

Independent Component Analysis (ICA) is a variable reduction technique that is used to reduce a large number of variables from a specific data set into a smaller number of artificial variables (independent components that will count for most of information contained in the variables). In our example, ten core interest rates from January 2001 to March 2015 are transformed into a smaller number of artificial variables. The non-Gaussian assumption of the underlying distribution of interest rates is needed in order for the method to work. It is different than the traditional principal component method, which require variable be normally distributed. In reality, most of financial time series are heavy-tailed, which means using ICA method is more appropriate than using PCA method. The linkage of extreme interest rate scenario and worst EVE via an assumption and the way to choose worst scenario will be discussed.

When: Wednesday, October 28, 2015 | 5:00 pm to 7:00 pm CDT
Where: University of Texas at Dallas
School of Management (SOM) Executive Education Dining Hall JSOM 1.502 800 West Campbell Road Richardson, Texas


Ke Chen
Model Validation Officer, Comerica Bank

For more information and to register for this chapter meeting, visit:
More Less


To improve this listing email:
NaN people listening