A Metaheuristic Approach for Classification Rule Discovery: AntMiner

In this transformational era of modern computer technology, information and knowledge is enhancing day by day and becoming a challenging task to manage, classify and mining the structured as well as unstructured data. Classification of discovered rules is an important activity responsible for future...

Full description

Saved in:
Bibliographic Details
Main Authors: Jyoti, Bhawna, Sharma, Aman Kumar
Format: Conference Proceeding
Language:English
Subjects:
Ant Colony System
antMiner
probability transition function
quality rule function
wilcoxon signed rank test
Online Access:Request full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:In this transformational era of modern computer technology, information and knowledge is enhancing day by day and becoming a challenging task to manage, classify and mining the structured as well as unstructured data. Classification of discovered rules is an important activity responsible for future predictions, extracting knowledge and patterns in the domain of data-mining. This paper deals with study of implementing nature-inspired (Ant Colony based technique) algorithm in classification rule generation in the field of data mining. The antMiner metaheuristic algorithm is applied on 12 standard datasets and simulation study is conducted(along with non-parametric statistical tests) which shows that there is significantly increase in accuracy measurement by c-antMiner2,c-antMinerpb over JRIP and PART algorithms.
ISSN:2640-074X
DOI:10.1109/ICIIP47207.2019.8985770