A multivariate neural forecasting modeling for air transport - Preprocessed by decomposition: A Brazilian application

被引：41

作者：

Alekseev, K. P. G. ^{[1
]}

Seixas, J. M. ^{[1
]}

机构：

[1] Univ Fed Rio de Janeiro, COPPE Poli, BR-21945970 Rio De Janeiro, Brazil

来源：

JOURNAL OF AIR TRANSPORT MANAGEMENT | 2009年 / 15卷 / 05期

关键词：

Neural networks; Time series; Air transport; Forecasting; Demand forecasting in air transport passenger; NETWORKS;

D O I：

10.1016/j.jairtraman.2008.08.008

中图分类号：

U [交通运输];

学科分类号：

08 ; 0823 ;

摘要：

An artificial neural forecasting model is developed for air transport passenger analysis. It uses a preprocessing method that decomposes information to reveal relevant features from the data. It is found that neural processing outperforms the traditional econometric approach and offers generalization on time series behavior, even where there are only small samples. (C) 2008 Elsevier Ltd. All rights reserved.

引用

页码：212 / 216

页数：5

共 18 条

[1] ALEKSEEV KCG, 2001, EURO 2001 EUR C OP R
[2] ALEKSEEV KCP, 1999, 5 INT C INF SCI AN S
[3] ALEKSEEV KCP, 2002, 22 INT S FOR
[4] [Anonymous], SSE EFI WORKING PAPE
[5] *CIV AV I, 1998, GLOB AIR TRANSP DEM
[6] DORSEY RE, 1998, FLEXIBLE FORM ESTIMA
[7] Haykin S., 1999, NEURAL NETWORKS COMP, DOI DOI 10.1017/S0269888998214044
[8] Neural network models for time series forecasts
Hill, T
OConnor, M
Remus, W
[J]. MANAGEMENT SCIENCE, 1996, 42 (07) : 1082 - 1092
[9] JURIK M, 1992, FUTURES, V21, P40
[10] Kolarik T., 1994, APL Quote Quad, V25, P86, DOI 10.1145/190468.190290

← 1 2 →