|
Title:
|
DATAMINING TOOL WITH EXPLORATORY SEARCH AND FEATURE DISCOVERY |
|
Author(s):
|
Hiroshi Sugimura, Kazunori Matsumoto |
|
ISBN:
|
978-972-8939-23-6 |
|
Editors:
|
António Palma dos Reis and Ajith P. Abraham |
|
Year:
|
2010 |
|
Edition:
|
Single |
|
Keywords:
|
data mining, machine learning, query, genetic algorithm, time-series data. |
|
Type:
|
Poster / Demonstration |
|
First Page:
|
147 |
|
Last Page:
|
150 |
|
Language:
|
English |
|
Cover:
|
|
|
Full Contents:
|
click to dowload 
|
|
Paper Abstract:
|
This paper proposes a system that datamines knowledge to extrapolate future behaviors of time-series data. This system includes two major tools. First one is a search tool which collects necessary information from databases. The standard database query language is extended to deal with typical properties of time-series data. We explain an overview of the language and a search mechanism. Second tool provides an automatic mechanism to discover features over a given set of training data. We can describe data in terms of discovered features. The discovery procedure begins with a set of randomly generated features and successively improved, generation by generation, using the genetic algorithm. By using the well optimized features we build a decision tree that predicts future behaviors. We explain how these two tools are combinatory applied in the entire knowledge discovery process. |
|
|
|
|
|
|
