The Effect of Normalization for Real Value Negative Selection Algorithm

Mohamad Farhan Mohamad Mohsin, Abdul Razak Hamdan, Azuraliza Abu Bakar

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

4 Citations (Scopus)

Abstract

The preliminary information of data being normalized into 0 and 1 is essential for an accurate data mining result including real value negative selection algorithm. As one class classification, only the self sample is available during normalization; therefore there is less confidence it fully represents the whole problem when the non-self sample is unknown. The problem 'out of range' arises when the values of data being monitored exceed the boundary as the setting in the normalizing phase. This study aimed to investigate the effect of normalization technique and identify the most reliable normalization algorithm for real value negative selection algorithm mainly when the non-self is not available. Three normalization algorithms - the min max, soft-max scaling, and z-scores were selected for the experiment. Four universal datasets were normalized and the performance of each normalization algorithm towards real value negative selection algorithm were measured based on five key performance metrics-detection rate, specificity, false alarm rate, accuracy, and number of detector. The result indicates that the real value negative selection is highly relied on type of normalization algorithm where the selection of appropriate normalization approach can improve detection performance. The min max is the most reliable algorithm for real value negative selection when it consistently produces a good detection performance. Similar to Z-score it also has similar capability however min max seems to a better approach in term of higher specificity, lower false alarm rate, and fewer numbers of detectors. Meanwhile, the soft max scaling is found not suitable for real value negative selection algorithm.

Original languageEnglish
Title of host publicationCommunications in Computer and Information Science
PublisherSpringer Verlag
Pages194-205
Number of pages12
Volume378 CCIS
ISBN (Print)9783642405662
DOIs
Publication statusPublished - 2013
Event2nd International Multi-Conference on Artificial Intelligence Technology, M-CAIT 2013 - Shah Alam
Duration: 28 Aug 201329 Aug 2013

Publication series

NameCommunications in Computer and Information Science
Volume378 CCIS
ISSN (Print)18650929

Other

Other2nd International Multi-Conference on Artificial Intelligence Technology, M-CAIT 2013
CityShah Alam
Period28/8/1329/8/13

Fingerprint

Detectors
Data mining
Experiments

Keywords

  • Min Max
  • Negative Selection Algorithm
  • Soft Scaling
  • Z-Score

ASJC Scopus subject areas

  • Computer Science(all)

Cite this

Mohamad Mohsin, M. F., Hamdan, A. R., & Abu Bakar, A. (2013). The Effect of Normalization for Real Value Negative Selection Algorithm. In Communications in Computer and Information Science (Vol. 378 CCIS, pp. 194-205). (Communications in Computer and Information Science; Vol. 378 CCIS). Springer Verlag. https://doi.org/10.1007/978-3-642-40567-9_17

The Effect of Normalization for Real Value Negative Selection Algorithm. / Mohamad Mohsin, Mohamad Farhan; Hamdan, Abdul Razak; Abu Bakar, Azuraliza.

Communications in Computer and Information Science. Vol. 378 CCIS Springer Verlag, 2013. p. 194-205 (Communications in Computer and Information Science; Vol. 378 CCIS).

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

Mohamad Mohsin, MF, Hamdan, AR & Abu Bakar, A 2013, The Effect of Normalization for Real Value Negative Selection Algorithm. in Communications in Computer and Information Science. vol. 378 CCIS, Communications in Computer and Information Science, vol. 378 CCIS, Springer Verlag, pp. 194-205, 2nd International Multi-Conference on Artificial Intelligence Technology, M-CAIT 2013, Shah Alam, 28/8/13. https://doi.org/10.1007/978-3-642-40567-9_17
Mohamad Mohsin MF, Hamdan AR, Abu Bakar A. The Effect of Normalization for Real Value Negative Selection Algorithm. In Communications in Computer and Information Science. Vol. 378 CCIS. Springer Verlag. 2013. p. 194-205. (Communications in Computer and Information Science). https://doi.org/10.1007/978-3-642-40567-9_17
Mohamad Mohsin, Mohamad Farhan ; Hamdan, Abdul Razak ; Abu Bakar, Azuraliza. / The Effect of Normalization for Real Value Negative Selection Algorithm. Communications in Computer and Information Science. Vol. 378 CCIS Springer Verlag, 2013. pp. 194-205 (Communications in Computer and Information Science).
@inproceedings{8650a89bba384a2c8c917d87815fd76c,
title = "The Effect of Normalization for Real Value Negative Selection Algorithm",
abstract = "The preliminary information of data being normalized into 0 and 1 is essential for an accurate data mining result including real value negative selection algorithm. As one class classification, only the self sample is available during normalization; therefore there is less confidence it fully represents the whole problem when the non-self sample is unknown. The problem 'out of range' arises when the values of data being monitored exceed the boundary as the setting in the normalizing phase. This study aimed to investigate the effect of normalization technique and identify the most reliable normalization algorithm for real value negative selection algorithm mainly when the non-self is not available. Three normalization algorithms - the min max, soft-max scaling, and z-scores were selected for the experiment. Four universal datasets were normalized and the performance of each normalization algorithm towards real value negative selection algorithm were measured based on five key performance metrics-detection rate, specificity, false alarm rate, accuracy, and number of detector. The result indicates that the real value negative selection is highly relied on type of normalization algorithm where the selection of appropriate normalization approach can improve detection performance. The min max is the most reliable algorithm for real value negative selection when it consistently produces a good detection performance. Similar to Z-score it also has similar capability however min max seems to a better approach in term of higher specificity, lower false alarm rate, and fewer numbers of detectors. Meanwhile, the soft max scaling is found not suitable for real value negative selection algorithm.",
keywords = "Min Max, Negative Selection Algorithm, Soft Scaling, Z-Score",
author = "{Mohamad Mohsin}, {Mohamad Farhan} and Hamdan, {Abdul Razak} and {Abu Bakar}, Azuraliza",
year = "2013",
doi = "10.1007/978-3-642-40567-9_17",
language = "English",
isbn = "9783642405662",
volume = "378 CCIS",
series = "Communications in Computer and Information Science",
publisher = "Springer Verlag",
pages = "194--205",
booktitle = "Communications in Computer and Information Science",

}

TY - GEN

T1 - The Effect of Normalization for Real Value Negative Selection Algorithm

AU - Mohamad Mohsin, Mohamad Farhan

AU - Hamdan, Abdul Razak

AU - Abu Bakar, Azuraliza

PY - 2013

Y1 - 2013

N2 - The preliminary information of data being normalized into 0 and 1 is essential for an accurate data mining result including real value negative selection algorithm. As one class classification, only the self sample is available during normalization; therefore there is less confidence it fully represents the whole problem when the non-self sample is unknown. The problem 'out of range' arises when the values of data being monitored exceed the boundary as the setting in the normalizing phase. This study aimed to investigate the effect of normalization technique and identify the most reliable normalization algorithm for real value negative selection algorithm mainly when the non-self is not available. Three normalization algorithms - the min max, soft-max scaling, and z-scores were selected for the experiment. Four universal datasets were normalized and the performance of each normalization algorithm towards real value negative selection algorithm were measured based on five key performance metrics-detection rate, specificity, false alarm rate, accuracy, and number of detector. The result indicates that the real value negative selection is highly relied on type of normalization algorithm where the selection of appropriate normalization approach can improve detection performance. The min max is the most reliable algorithm for real value negative selection when it consistently produces a good detection performance. Similar to Z-score it also has similar capability however min max seems to a better approach in term of higher specificity, lower false alarm rate, and fewer numbers of detectors. Meanwhile, the soft max scaling is found not suitable for real value negative selection algorithm.

AB - The preliminary information of data being normalized into 0 and 1 is essential for an accurate data mining result including real value negative selection algorithm. As one class classification, only the self sample is available during normalization; therefore there is less confidence it fully represents the whole problem when the non-self sample is unknown. The problem 'out of range' arises when the values of data being monitored exceed the boundary as the setting in the normalizing phase. This study aimed to investigate the effect of normalization technique and identify the most reliable normalization algorithm for real value negative selection algorithm mainly when the non-self is not available. Three normalization algorithms - the min max, soft-max scaling, and z-scores were selected for the experiment. Four universal datasets were normalized and the performance of each normalization algorithm towards real value negative selection algorithm were measured based on five key performance metrics-detection rate, specificity, false alarm rate, accuracy, and number of detector. The result indicates that the real value negative selection is highly relied on type of normalization algorithm where the selection of appropriate normalization approach can improve detection performance. The min max is the most reliable algorithm for real value negative selection when it consistently produces a good detection performance. Similar to Z-score it also has similar capability however min max seems to a better approach in term of higher specificity, lower false alarm rate, and fewer numbers of detectors. Meanwhile, the soft max scaling is found not suitable for real value negative selection algorithm.

KW - Min Max

KW - Negative Selection Algorithm

KW - Soft Scaling

KW - Z-Score

UR - http://www.scopus.com/inward/record.url?scp=84904678151&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84904678151&partnerID=8YFLogxK

U2 - 10.1007/978-3-642-40567-9_17

DO - 10.1007/978-3-642-40567-9_17

M3 - Conference contribution

AN - SCOPUS:84904678151

SN - 9783642405662

VL - 378 CCIS

T3 - Communications in Computer and Information Science

SP - 194

EP - 205

BT - Communications in Computer and Information Science

PB - Springer Verlag

ER -