RT Journal Article
ID 53ca3faa6fc50c16
A1 Fedorov, Andrey V.
A1 Bidyuk, Petr I.
T1 Comparative Analysis of Some Forecasting Methods on Nonstationary Processes
JF Journal of Automation and Information Sciences
JO JAI(S)
YR 2008
FD 2008-05-22
VO 40
IS 3
SP 59
OP 69
AB Two types of mathematical models are proposed to forecast the process of stock price formation on gold. To ensure the insightful analysis of the nonstationary processes of this kind, the probabilistic dynamic Bayesian network is proposed. GMDH and autoregressive model that supplemented the probabilistic model improved the quality of decisions taken while stock exchange operations. The comparative analysis is performed and the best models are selected for the processes analyzed. The model for a short-term forecast of the conditional variance of the process is constructed.
PB Begell House
LK https://www.dl.begellhouse.com/journals/2b6239406278e43e,3b3ce9c8499e64af,53ca3faa6fc50c16.html