Using Significant Classification Rules to Analyze Korean Customers´ Power Consumption Behavior: Incremental Tree Induction using Cascading-and-Sharing Method

Author

Piao, Minghao ; Li, Meijing ; Ryu, Keun Ho

Author_Institution

Database/Bio Inf. Lab., Chungbuk Nat. Univ., Cheongju, South Korea

fYear

2010

fDate

June 29 2010-July 1 2010

Firstpage

1649

Lastpage

1653

Abstract

Power load analysis is an important issue in electrical industry. Data mining techniques are widely studied methodology for power load analysis and it helps decision making on electrical industry. In this paper, we propose an incremental tree induction algorithm using Cascading-and-Sharing method, and use mined significant classification rules to analyze customers´ power consumption behavior in General, Education and Regular groups.

Keywords

data mining; decision making; decision trees; electricity supply industry; load (electric); pattern classification; power consumption; power engineering computing; Korean customers power consumption behavior analysis; cascading-and-sharing method; data mining technique; decision making; electrical industry; incremental tree induction algorithm; power load analysis; significant classification rules; Classification algorithms; Classification tree analysis; Data mining; Load forecasting; Temperature sensors; Training; decision tree induction; incremental mining; power consumption behavior; significant classification rules;

fLanguage

English

Publisher

ieee

Conference_Titel

Computer and Information Technology (CIT), 2010 IEEE 10th International Conference on

Conference_Location

Bradford

Print_ISBN

978-1-4244-7547-6

Type

conf

DOI

10.1109/CIT.2010.503

Filename

5577952