A modal symbolic classifier for selecting time series models

被引：20

作者：

Prudêncio, RBC ^{[1
]}

Ludermir, TB ^{[1
]}

de Carvalho, FDT ^{[1
]}

机构：

[1] Univ Fed Pernambuco, Ctr Informat, BR-50732970 Recife, PE, Brazil

来源：

PATTERN RECOGNITION LETTERS | 2004年 / 25卷 / 08期

关键词：

time series forecasting; model selection; machine learning; symbolic data analysis; modal symbolic classifier;

D O I：

10.1016/j.patrec.2004.02.004

中图分类号：

TP18 [人工智能理论];

学科分类号：

081104 ; 0812 ; 0835 ; 1405 ;

摘要：

The selection of a good model for forecasting a time series is a task that involves experience and knowledge. Employing machine learning algorithms is a promising approach to acquiring knowledge in regards to this task. A supervised classification method originating from the symbolic data analysis field is proposed for the model selection problem. This method was applied in the task of selecting between two widespread models, and compared to other learning algorithms. To date, it has obtained the lowest classification errors among all the tested algorithms. (C) 2004 Elsevier B.V. All rights reserved.

引用

页码：911 / 921

页数：11

共 28 条

[1]

[Anonymous], 1992, EXPERT SYSTEMS APPL

[2]

[Anonymous], 1993, P 13 INT JOINT C ART

[3] SELECTING APPROPRIATE FORECASTING MODELS USING RULE INDUCTION [J].

ARINZE, B .

OMEGA-INTERNATIONAL JOURNAL OF MANAGEMENT SCIENCE, 1994, 22 (06) :647-658

[4]

Bezerra B, 2002, LECT NOTES ARTIF INT, V2507, P227

[5]

Bock HH, 2000, ST CLASS DAT ANAL, P39

[6]

Bock HH., 2000, ANAL SYMBOLIC DATA

[7]

Box G. E. P, 1970, TIME SERIES ANAL FOR

[8]

Brown R. G., 1963, SMOOTHING FORECASTIN

[9] NEURAL-NETWORK SYSTEM FOR FORECASTING METHOD SELECTION [J].

CHU, CH ;

WIDJAJA, D .

DECISION SUPPORT SYSTEMS, 1994, 12 (01) :13-24

[10] RULE-BASED FORECASTING - DEVELOPMENT AND VALIDATION OF AN EXPERT SYSTEMS-APPROACH TO COMBINING TIME-SERIES EXTRAPOLATIONS [J].

COLLOPY, F ;

ARMSTRONG, JS .

MANAGEMENT SCIENCE, 1992, 38 (10) :1394-1414

← 1 2 3 →