2. Kernel Estimation Flashcards
What is Kernel estimation about ?
What are some examples of Kernels and what can they tell about the underlying distribution?
What is the role of h (bandwidth) in kernel estimation?
What is the intuition behind the rule of thumb?
Is choice of kernel or bandwidth choice more important?
How do you estimate conditional kernels?
How do you estimate bivariate densities and what do you need to careful with when doing so?
What does large excess kurtosis reveal about the choice of distribution?
What should you keep in mind when presented a histogram?
What are some extensions to kernel estimation ? What are the problems with them and their solutions?
What are the different types of loss given defaults? Who are interested in each?
What is the hiererchy of payback with defaults and what can you expect looking at graphs depicting payback of them?
Discuss how beta estimation vs kernel estimation performs regarding VaR and its implications
