Forecasting Counting Processes by PCP Models

A.M. Aguilera, P.R. Bouzas, and M.J. Valderrama (Spain)

Keywords

Modelling; Functional data; Principal component; Poissonprocess.

Abstract

A Principal Component Prediction (PCP) model for count ing processes is derived in this paper from a weighted esti mation of the functional PCA involved. An application to a queue model is also included.

Important Links:

• DOI:
• From Proceeding (350) Modelling, Identification, and Control - 2002

Go Back