Functional Data Analysis of Payment Systems

In this paper for a credit cards payment system as robust predictor of transactions number and transactions intensity is proposed by means of functional autoregressive model. Intraday economic time series are treated as random continuous functions projected onto low dimensional subspace. Both B-spl...

Full description

Bibliographic Details
Main Authors: A. Laukaitis, A. Račkauskas
Format: Article
Language:English
Published: Vilnius University Press 2002-12-01
Series:Nonlinear Analysis
Subjects:
Online Access:http://www.zurnalai.vu.lt/nonlinear-analysis/article/view/15194
Description
Summary:In this paper for a credit cards payment system as robust predictor of transactions number and transactions intensity is proposed by means of functional autoregressive model. Intraday economic time series are treated as random continuous functions projected onto low dimensional subspace. Both B-splines and Fourier bases are considered for data smoothing.
ISSN:1392-5113
2335-8963