Summary
2011 International Symposium on Nonlinear Theory and Its Applications
2011
Session Number:B2L-B
Session:
Number:B2L-B2
Estimation of Transition of Credit Rating by using Particle Filters based on State Equations approximated by the Genetic Programming
Shozo Tokinaga, Seigo Matsuno,
pp.397-400
Publication Date:2011/9/4
Online ISSN:2188-5079
Summary:
This paper deals with the estimation of transition of credit rating by using Particle Filters (PFs) based on state equations approximated by the Genetic Programming (GP). Our aim is to find a true rating from observed ratings usually corrupted by noise, however, these works include simple scheme of time series modeling. In this paper, we generalize the PFs so that we approximate state equations by using the GP where individuals corresponding to state equations are improved according to the likelihood of PFs. At the same time, we include dynamics of financial ratios besides ratings in the nonlinear system equations, then we can expect improved estimation of true ratings.