Classification with missing data using multi-layered artificial immune systems

Mlungisi Duma; Bhekisipho Twala; Tshilidzi Marwala; Fulufhelo Vincent Nelwamondo

doi:10.1109/CEC.2012.6256420

Classification with missing data using multi-layered artificial immune systems

Mlungisi Duma, Bhekisipho Twala, Tshilidzi Marwala, Fulufhelo Vincent Nelwamondo

Faculty of Engineering and the Built Environment

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

4 Citations (Scopus)

Abstract

The nature of missing data problems forces us to build models that maintain high accuracies and steadiness. The models developed to achieve this are usually complex and computationally expensive. In this paper, we propose an unsupervised multi-layered artificial immune system for an insurance classification problem that is characterised as highly dimensional and contains escalating missing data. The system is compared with the k-nearest neighbour, support vector machines and logistic discriminant models. Overall, the results show that whilst k-nearest neighbour achieves the highest accuracy, the multi-layered artificial immune system is steady and maintains high performance compared to other models, regardless of how the missing data is distributed in a dataset.

Original language	English
Title of host publication	2012 IEEE Congress on Evolutionary Computation, CEC 2012
DOIs	https://doi.org/10.1109/CEC.2012.6256420
Publication status	Published - 2012
Event	2012 IEEE Congress on Evolutionary Computation, CEC 2012 - Brisbane, QLD, Australia Duration: 10 Jun 2012 → 15 Jun 2012

Publication series

Name	2012 IEEE Congress on Evolutionary Computation, CEC 2012

Conference

Conference	2012 IEEE Congress on Evolutionary Computation, CEC 2012
Country/Territory	Australia
City	Brisbane, QLD
Period	10/06/12 → 15/06/12

Keywords

insurance risk classification
missing dat
multi-layered artificial immune system

ASJC Scopus subject areas

Computational Theory and Mathematics
Theoretical Computer Science

Access to Document

10.1109/CEC.2012.6256420

Cite this

@inproceedings{8896ad5aba5d418bb1aea0c70cf89516,

title = "Classification with missing data using multi-layered artificial immune systems",

abstract = "The nature of missing data problems forces us to build models that maintain high accuracies and steadiness. The models developed to achieve this are usually complex and computationally expensive. In this paper, we propose an unsupervised multi-layered artificial immune system for an insurance classification problem that is characterised as highly dimensional and contains escalating missing data. The system is compared with the k-nearest neighbour, support vector machines and logistic discriminant models. Overall, the results show that whilst k-nearest neighbour achieves the highest accuracy, the multi-layered artificial immune system is steady and maintains high performance compared to other models, regardless of how the missing data is distributed in a dataset.",

keywords = "insurance risk classification, missing dat, multi-layered artificial immune system",

author = "Mlungisi Duma and Bhekisipho Twala and Tshilidzi Marwala and Nelwamondo, {Fulufhelo Vincent}",

year = "2012",

doi = "10.1109/CEC.2012.6256420",

language = "English",

isbn = "9781467315098",

series = "2012 IEEE Congress on Evolutionary Computation, CEC 2012",

booktitle = "2012 IEEE Congress on Evolutionary Computation, CEC 2012",

note = "2012 IEEE Congress on Evolutionary Computation, CEC 2012 ; Conference date: 10-06-2012 Through 15-06-2012",

}

Duma, M, Twala, B, Marwala, T & Nelwamondo, FV 2012, Classification with missing data using multi-layered artificial immune systems. in 2012 IEEE Congress on Evolutionary Computation, CEC 2012., 6256420, 2012 IEEE Congress on Evolutionary Computation, CEC 2012, 2012 IEEE Congress on Evolutionary Computation, CEC 2012, Brisbane, QLD, Australia, 10/06/12. https://doi.org/10.1109/CEC.2012.6256420

Classification with missing data using multi-layered artificial immune systems. / Duma, Mlungisi; Twala, Bhekisipho; Marwala, Tshilidzi et al.
2012 IEEE Congress on Evolutionary Computation, CEC 2012. 2012. 6256420 (2012 IEEE Congress on Evolutionary Computation, CEC 2012).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

TY - GEN

T1 - Classification with missing data using multi-layered artificial immune systems

AU - Duma, Mlungisi

AU - Twala, Bhekisipho

AU - Marwala, Tshilidzi

AU - Nelwamondo, Fulufhelo Vincent

PY - 2012

Y1 - 2012

N2 - The nature of missing data problems forces us to build models that maintain high accuracies and steadiness. The models developed to achieve this are usually complex and computationally expensive. In this paper, we propose an unsupervised multi-layered artificial immune system for an insurance classification problem that is characterised as highly dimensional and contains escalating missing data. The system is compared with the k-nearest neighbour, support vector machines and logistic discriminant models. Overall, the results show that whilst k-nearest neighbour achieves the highest accuracy, the multi-layered artificial immune system is steady and maintains high performance compared to other models, regardless of how the missing data is distributed in a dataset.

AB - The nature of missing data problems forces us to build models that maintain high accuracies and steadiness. The models developed to achieve this are usually complex and computationally expensive. In this paper, we propose an unsupervised multi-layered artificial immune system for an insurance classification problem that is characterised as highly dimensional and contains escalating missing data. The system is compared with the k-nearest neighbour, support vector machines and logistic discriminant models. Overall, the results show that whilst k-nearest neighbour achieves the highest accuracy, the multi-layered artificial immune system is steady and maintains high performance compared to other models, regardless of how the missing data is distributed in a dataset.

KW - insurance risk classification

KW - missing dat

KW - multi-layered artificial immune system

UR - http://www.scopus.com/inward/record.url?scp=84866865993&partnerID=8YFLogxK

U2 - 10.1109/CEC.2012.6256420

DO - 10.1109/CEC.2012.6256420

M3 - Conference contribution

AN - SCOPUS:84866865993

SN - 9781467315098

T3 - 2012 IEEE Congress on Evolutionary Computation, CEC 2012

BT - 2012 IEEE Congress on Evolutionary Computation, CEC 2012

T2 - 2012 IEEE Congress on Evolutionary Computation, CEC 2012

Y2 - 10 June 2012 through 15 June 2012

ER -

Classification with missing data using multi-layered artificial immune systems

Abstract

Publication series

Conference

Keywords

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this