PERFICT: Perturbed frequent itemset based classification technique

RaghvenPhDa Mall, Prakhar Jain, Vikram Pudi

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

This paper presents Perturbed Frequent Itemset based Classification Technique (PERFICT), a novel associative classification approach based on perturbed frequent itemsets. Most of the existing associative classifiers work well on transactional data where each record contains a set of boolean items. They are not very effective in general for relational data that typically contains real valued attributes. In PERFICT, we handle real attributes by treating items as (attribute,value) pairs, where the value is not the original one, but is perturbed by a small amount and is a range based value. We also propose our own similarity measure which captures the nature of real valued attributes and provide effective weights for the itemsets. The probabilistic contributions of different itemsets is taken into considerations during classification. Some of the applications where such a technique is useful are in signal classification, medical diagnosis and handwriting recognition. Experiments conducted on the UCI Repository datasets show that PERFICT is highly competitive in terms of accuracy in comparison with popular associative classification methods.

Original languageEnglish
Title of host publicationProceedings - 22nd International Conference on Tools with Artificial Intelligence, ICTAI 2010
Pages79-86
Number of pages8
Volume1
DOIs
Publication statusPublished - 1 Dec 2010
Externally publishedYes
Event22nd International Conference on Tools with Artificial Intelligence, ICTAI 2010 - Arras, France
Duration: 27 Oct 201029 Oct 2010

Other

Other22nd International Conference on Tools with Artificial Intelligence, ICTAI 2010
CountryFrance
CityArras
Period27/10/1029/10/10

Fingerprint

Classifiers
Experiments

ASJC Scopus subject areas

  • Software
  • Artificial Intelligence
  • Computer Science Applications

Cite this

Mall, R., Jain, P., & Pudi, V. (2010). PERFICT: Perturbed frequent itemset based classification technique. In Proceedings - 22nd International Conference on Tools with Artificial Intelligence, ICTAI 2010 (Vol. 1, pp. 79-86). [5670019] https://doi.org/10.1109/ICTAI.2010.20

PERFICT : Perturbed frequent itemset based classification technique. / Mall, RaghvenPhDa; Jain, Prakhar; Pudi, Vikram.

Proceedings - 22nd International Conference on Tools with Artificial Intelligence, ICTAI 2010. Vol. 1 2010. p. 79-86 5670019.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Mall, R, Jain, P & Pudi, V 2010, PERFICT: Perturbed frequent itemset based classification technique. in Proceedings - 22nd International Conference on Tools with Artificial Intelligence, ICTAI 2010. vol. 1, 5670019, pp. 79-86, 22nd International Conference on Tools with Artificial Intelligence, ICTAI 2010, Arras, France, 27/10/10. https://doi.org/10.1109/ICTAI.2010.20
Mall R, Jain P, Pudi V. PERFICT: Perturbed frequent itemset based classification technique. In Proceedings - 22nd International Conference on Tools with Artificial Intelligence, ICTAI 2010. Vol. 1. 2010. p. 79-86. 5670019 https://doi.org/10.1109/ICTAI.2010.20
Mall, RaghvenPhDa ; Jain, Prakhar ; Pudi, Vikram. / PERFICT : Perturbed frequent itemset based classification technique. Proceedings - 22nd International Conference on Tools with Artificial Intelligence, ICTAI 2010. Vol. 1 2010. pp. 79-86
@inproceedings{5ca20c3bcfb847959e55ac5e50cd78cc,
title = "PERFICT: Perturbed frequent itemset based classification technique",
abstract = "This paper presents Perturbed Frequent Itemset based Classification Technique (PERFICT), a novel associative classification approach based on perturbed frequent itemsets. Most of the existing associative classifiers work well on transactional data where each record contains a set of boolean items. They are not very effective in general for relational data that typically contains real valued attributes. In PERFICT, we handle real attributes by treating items as (attribute,value) pairs, where the value is not the original one, but is perturbed by a small amount and is a range based value. We also propose our own similarity measure which captures the nature of real valued attributes and provide effective weights for the itemsets. The probabilistic contributions of different itemsets is taken into considerations during classification. Some of the applications where such a technique is useful are in signal classification, medical diagnosis and handwriting recognition. Experiments conducted on the UCI Repository datasets show that PERFICT is highly competitive in terms of accuracy in comparison with popular associative classification methods.",
author = "RaghvenPhDa Mall and Prakhar Jain and Vikram Pudi",
year = "2010",
month = "12",
day = "1",
doi = "10.1109/ICTAI.2010.20",
language = "English",
isbn = "9780769542638",
volume = "1",
pages = "79--86",
booktitle = "Proceedings - 22nd International Conference on Tools with Artificial Intelligence, ICTAI 2010",

}

TY - GEN

T1 - PERFICT

T2 - Perturbed frequent itemset based classification technique

AU - Mall, RaghvenPhDa

AU - Jain, Prakhar

AU - Pudi, Vikram

PY - 2010/12/1

Y1 - 2010/12/1

N2 - This paper presents Perturbed Frequent Itemset based Classification Technique (PERFICT), a novel associative classification approach based on perturbed frequent itemsets. Most of the existing associative classifiers work well on transactional data where each record contains a set of boolean items. They are not very effective in general for relational data that typically contains real valued attributes. In PERFICT, we handle real attributes by treating items as (attribute,value) pairs, where the value is not the original one, but is perturbed by a small amount and is a range based value. We also propose our own similarity measure which captures the nature of real valued attributes and provide effective weights for the itemsets. The probabilistic contributions of different itemsets is taken into considerations during classification. Some of the applications where such a technique is useful are in signal classification, medical diagnosis and handwriting recognition. Experiments conducted on the UCI Repository datasets show that PERFICT is highly competitive in terms of accuracy in comparison with popular associative classification methods.

AB - This paper presents Perturbed Frequent Itemset based Classification Technique (PERFICT), a novel associative classification approach based on perturbed frequent itemsets. Most of the existing associative classifiers work well on transactional data where each record contains a set of boolean items. They are not very effective in general for relational data that typically contains real valued attributes. In PERFICT, we handle real attributes by treating items as (attribute,value) pairs, where the value is not the original one, but is perturbed by a small amount and is a range based value. We also propose our own similarity measure which captures the nature of real valued attributes and provide effective weights for the itemsets. The probabilistic contributions of different itemsets is taken into considerations during classification. Some of the applications where such a technique is useful are in signal classification, medical diagnosis and handwriting recognition. Experiments conducted on the UCI Repository datasets show that PERFICT is highly competitive in terms of accuracy in comparison with popular associative classification methods.

UR - http://www.scopus.com/inward/record.url?scp=78751533733&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=78751533733&partnerID=8YFLogxK

U2 - 10.1109/ICTAI.2010.20

DO - 10.1109/ICTAI.2010.20

M3 - Conference contribution

AN - SCOPUS:78751533733

SN - 9780769542638

VL - 1

SP - 79

EP - 86

BT - Proceedings - 22nd International Conference on Tools with Artificial Intelligence, ICTAI 2010

ER -