ENNA: Software effort estimation using ensemble of neural networks with associative memory

Yigit Kultur; Burak Turhan; Ayse Basar Bener

doi:10.1145/1453101.1453148

ENNA: Software effort estimation using ensemble of neural networks with associative memory

Yigit Kultur
, Burak Turhan
, Ayse Basar Bener

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

27 Citations (Scopus)

Abstract

Companies usually have limited amount of data for effort estimation. Machine learning methods have been preferred over parametric models due to their flexibility to calibrate the model for the available data. On the other hand, as machine learning methods become more complex they need more data to learn from. Therefore the challenge is to increase the performance of the algorithm when there is limited data. In this research we used a relatively complex machine learning algorithm, neural networks, and showed that stable and accurate estimations are achievable with an ensemble using associative memory. Our experimental results revealed that our proposed algorithm (ENNA) achieves on the average PRED(25) = 36.4 which is a significant increase compared to Neural Network (NN) PRED(25) = 8.

Original language	English
Title of host publication	SIGSOFT 2008/FSE-16 - Proceedings of the 16th ACM SIGSOFT International Symposium on the Foundations of Software Engineering
Pages	330-338
Number of pages	9
DOIs	https://doi.org/10.1145/1453101.1453148
Publication status	Published - 1 Dec 2008
Externally published	Yes
Event	ACM SIGSOFT International Symposium on the Foundations of Software Engineering 2008 - Atlanta, GA, United States of America Duration: 9 Nov 2008 → 14 Nov 2008 Conference number: 16th

Conference

Conference	ACM SIGSOFT International Symposium on the Foundations of Software Engineering 2008
Abbreviated title	SIGSOFT 2008/FSE-16
Country/Territory	United States of America
City	Atlanta, GA
Period	9/11/08 → 14/11/08

Keywords

Adaptive Resonance Theory
Associative Memory
Bootstrap
Cost Estimation
Effort Estimation
Ensemble
K nearest neighbors
Multilayer Perceptron
Neural Network

Access to Document

10.1145/1453101.1453148

Cite this

@inproceedings{de41900064514e55ab29278821e33254,

title = "ENNA: Software effort estimation using ensemble of neural networks with associative memory",

abstract = "Companies usually have limited amount of data for effort estimation. Machine learning methods have been preferred over parametric models due to their flexibility to calibrate the model for the available data. On the other hand, as machine learning methods become more complex they need more data to learn from. Therefore the challenge is to increase the performance of the algorithm when there is limited data. In this research we used a relatively complex machine learning algorithm, neural networks, and showed that stable and accurate estimations are achievable with an ensemble using associative memory. Our experimental results revealed that our proposed algorithm (ENNA) achieves on the average PRED(25) = 36.4 which is a significant increase compared to Neural Network (NN) PRED(25) = 8.",

keywords = "Adaptive Resonance Theory, Associative Memory, Bootstrap, Cost Estimation, Effort Estimation, Ensemble, K nearest neighbors, Multilayer Perceptron, Neural Network",

author = "Yigit Kultur and Burak Turhan and Bener, \{Ayse Basar\}",

year = "2008",

month = dec,

day = "1",

doi = "10.1145/1453101.1453148",

language = "English",

isbn = "9781595939951",

pages = "330--338",

booktitle = "SIGSOFT 2008/FSE-16 - Proceedings of the 16th ACM SIGSOFT International Symposium on the Foundations of Software Engineering",

note = "ACM SIGSOFT International Symposium on the Foundations of Software Engineering 2008, SIGSOFT 2008/FSE-16 ; Conference date: 09-11-2008 Through 14-11-2008",

}

Kultur, Y, Turhan, B & Bener, AB 2008, ENNA: Software effort estimation using ensemble of neural networks with associative memory. in SIGSOFT 2008/FSE-16 - Proceedings of the 16th ACM SIGSOFT International Symposium on the Foundations of Software Engineering. pp. 330-338, ACM SIGSOFT International Symposium on the Foundations of Software Engineering 2008, Atlanta, GA, United States of America, 9/11/08. https://doi.org/10.1145/1453101.1453148

ENNA: Software effort estimation using ensemble of neural networks with associative memory. / Kultur, Yigit; Turhan, Burak; Bener, Ayse Basar.
SIGSOFT 2008/FSE-16 - Proceedings of the 16th ACM SIGSOFT International Symposium on the Foundations of Software Engineering. 2008. p. 330-338.

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

TY - GEN

T1 - ENNA

T2 - ACM SIGSOFT International Symposium on the Foundations of Software Engineering 2008

AU - Kultur, Yigit

AU - Turhan, Burak

AU - Bener, Ayse Basar

N1 - Conference code: 16th

PY - 2008/12/1

Y1 - 2008/12/1

N2 - Companies usually have limited amount of data for effort estimation. Machine learning methods have been preferred over parametric models due to their flexibility to calibrate the model for the available data. On the other hand, as machine learning methods become more complex they need more data to learn from. Therefore the challenge is to increase the performance of the algorithm when there is limited data. In this research we used a relatively complex machine learning algorithm, neural networks, and showed that stable and accurate estimations are achievable with an ensemble using associative memory. Our experimental results revealed that our proposed algorithm (ENNA) achieves on the average PRED(25) = 36.4 which is a significant increase compared to Neural Network (NN) PRED(25) = 8.

AB - Companies usually have limited amount of data for effort estimation. Machine learning methods have been preferred over parametric models due to their flexibility to calibrate the model for the available data. On the other hand, as machine learning methods become more complex they need more data to learn from. Therefore the challenge is to increase the performance of the algorithm when there is limited data. In this research we used a relatively complex machine learning algorithm, neural networks, and showed that stable and accurate estimations are achievable with an ensemble using associative memory. Our experimental results revealed that our proposed algorithm (ENNA) achieves on the average PRED(25) = 36.4 which is a significant increase compared to Neural Network (NN) PRED(25) = 8.

KW - Adaptive Resonance Theory

KW - Associative Memory

KW - Bootstrap

KW - Cost Estimation

KW - Effort Estimation

KW - Ensemble

KW - K nearest neighbors

KW - Multilayer Perceptron

KW - Neural Network

UR - https://www.scopus.com/pages/publications/73949114934

U2 - 10.1145/1453101.1453148

DO - 10.1145/1453101.1453148

M3 - Conference Paper

AN - SCOPUS:73949114934

SN - 9781595939951

SP - 330

EP - 338

BT - SIGSOFT 2008/FSE-16 - Proceedings of the 16th ACM SIGSOFT International Symposium on the Foundations of Software Engineering

Y2 - 9 November 2008 through 14 November 2008

ER -

ENNA: Software effort estimation using ensemble of neural networks with associative memory

Abstract

Conference

Keywords

Access to Document

Other files and links

Cite this