Missing value imputation on multiple measurements for prediction of liver cancer recurrence: A comparative study

Xiao Ou Ping; Yi Ju Tseng; Ja Der Liang; Guan Tarn Huang; Pei Ming Yang; Feipei Lai

doi:10.3233/978-1-61499-484-8-1930

Missing value imputation on multiple measurements for prediction of liver cancer recurrence: A comparative study

Xiao Ou Ping^*, Yi Ju Tseng, Ja Der Liang, Guan Tarn Huang, Pei Ming Yang, Feipei Lai

^*Corresponding author for this work

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

Abstract

The problem of missing values frequently occurs during data analysis. Imputation is one of the solutions to handle missing data. Clinical data often contain multiple measurements such as laboratory test results which are measured at different time points. In this study, we compared three imputation methods and their effects on different multiple measurement data sets with different sampling time periods. Data sets of liver cancer were used in this study for classification of liver cancer recurrence based on two types of classification models built by support vector machine (SVM) and random forests. The results report appropriate combinations of imputation methods and sampling time periods which achieve better classification results than those of other imputation methods and periods. These reported the leading imputation method with SVM is significantly different (P<0.001) from mean imputation with SVM which is frequently used by data sets with missing values.

Original language	English
Title of host publication	Intelligent Systems and Applications - Proceedings of the International Computer Symposium, ICS 2014
Editors	William Cheng-Chung Chu, Han-Chieh Chao, Stephen Jenn-Hwa Yang
Publisher	IOS Press BV
Pages	1930-1939
Number of pages	10
ISBN (Electronic)	9781614994831
DOIs	https://doi.org/10.3233/978-1-61499-484-8-1930
State	Published - 2015
Externally published	Yes
Event	International Computer Symposium, ICS 2014 - Taichung, Taiwan Duration: 12 12 2014 → 14 12 2014

Publication series

Name	Frontiers in Artificial Intelligence and Applications
Volume	274
ISSN (Print)	0922-6389
ISSN (Electronic)	1879-8314

Conference

Conference	International Computer Symposium, ICS 2014
Country/Territory	Taiwan
City	Taichung
Period	12/12/14 → 14/12/14

Bibliographical note

Publisher Copyright:
© 2015 The authors and IOS Press. All rights reserved.

Keywords

Missing values
imputation
random forests
support vector machine

UN SDGs

This output contributes to the following UN Sustainable Development Goals (SDGs)

Access to Document

10.3233/978-1-61499-484-8-1930

Cite this

Ping, X. O., Tseng, Y. J., Liang, J. D., Huang, G. T., Yang, P. M., & Lai, F. (2015). Missing value imputation on multiple measurements for prediction of liver cancer recurrence: A comparative study. In W. C.-C. Chu, H.-C. Chao, & S. J.-H. Yang (Eds.), Intelligent Systems and Applications - Proceedings of the International Computer Symposium, ICS 2014 (pp. 1930-1939). (Frontiers in Artificial Intelligence and Applications; Vol. 274). IOS Press BV. https://doi.org/10.3233/978-1-61499-484-8-1930

Ping, Xiao Ou ; Tseng, Yi Ju ; Liang, Ja Der et al. / Missing value imputation on multiple measurements for prediction of liver cancer recurrence : A comparative study. Intelligent Systems and Applications - Proceedings of the International Computer Symposium, ICS 2014. editor / William Cheng-Chung Chu ; Han-Chieh Chao ; Stephen Jenn-Hwa Yang. IOS Press BV, 2015. pp. 1930-1939 (Frontiers in Artificial Intelligence and Applications).

@inproceedings{d856529451914f7f8c86b2d643e8a36c,

title = "Missing value imputation on multiple measurements for prediction of liver cancer recurrence: A comparative study",

abstract = "The problem of missing values frequently occurs during data analysis. Imputation is one of the solutions to handle missing data. Clinical data often contain multiple measurements such as laboratory test results which are measured at different time points. In this study, we compared three imputation methods and their effects on different multiple measurement data sets with different sampling time periods. Data sets of liver cancer were used in this study for classification of liver cancer recurrence based on two types of classification models built by support vector machine (SVM) and random forests. The results report appropriate combinations of imputation methods and sampling time periods which achieve better classification results than those of other imputation methods and periods. These reported the leading imputation method with SVM is significantly different (P<0.001) from mean imputation with SVM which is frequently used by data sets with missing values.",

keywords = "Missing values, imputation, random forests, support vector machine",

author = "Ping, {Xiao Ou} and Tseng, {Yi Ju} and Liang, {Ja Der} and Huang, {Guan Tarn} and Yang, {Pei Ming} and Feipei Lai",

year = "2015",

doi = "10.3233/978-1-61499-484-8-1930",

language = "英语",

series = "Frontiers in Artificial Intelligence and Applications",

publisher = "IOS Press BV",

pages = "1930--1939",

editor = "Chu, {William Cheng-Chung} and Han-Chieh Chao and Yang, {Stephen Jenn-Hwa}",

booktitle = "Intelligent Systems and Applications - Proceedings of the International Computer Symposium, ICS 2014",

address = "荷兰",

}

Ping, XO, Tseng, YJ, Liang, JD, Huang, GT, Yang, PM & Lai, F 2015, Missing value imputation on multiple measurements for prediction of liver cancer recurrence: A comparative study. in WC-C Chu, H-C Chao & SJ-H Yang (eds), Intelligent Systems and Applications - Proceedings of the International Computer Symposium, ICS 2014. Frontiers in Artificial Intelligence and Applications, vol. 274, IOS Press BV, pp. 1930-1939, International Computer Symposium, ICS 2014, Taichung, Taiwan, 12/12/14. https://doi.org/10.3233/978-1-61499-484-8-1930

Missing value imputation on multiple measurements for prediction of liver cancer recurrence: A comparative study. / Ping, Xiao Ou; Tseng, Yi Ju; Liang, Ja Der et al.
Intelligent Systems and Applications - Proceedings of the International Computer Symposium, ICS 2014. ed. / William Cheng-Chung Chu; Han-Chieh Chao; Stephen Jenn-Hwa Yang. IOS Press BV, 2015. p. 1930-1939 (Frontiers in Artificial Intelligence and Applications; Vol. 274).

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

TY - GEN

T1 - Missing value imputation on multiple measurements for prediction of liver cancer recurrence

T2 - International Computer Symposium, ICS 2014

AU - Ping, Xiao Ou

AU - Tseng, Yi Ju

AU - Liang, Ja Der

AU - Huang, Guan Tarn

AU - Yang, Pei Ming

AU - Lai, Feipei

PY - 2015

Y1 - 2015

N2 - The problem of missing values frequently occurs during data analysis. Imputation is one of the solutions to handle missing data. Clinical data often contain multiple measurements such as laboratory test results which are measured at different time points. In this study, we compared three imputation methods and their effects on different multiple measurement data sets with different sampling time periods. Data sets of liver cancer were used in this study for classification of liver cancer recurrence based on two types of classification models built by support vector machine (SVM) and random forests. The results report appropriate combinations of imputation methods and sampling time periods which achieve better classification results than those of other imputation methods and periods. These reported the leading imputation method with SVM is significantly different (P<0.001) from mean imputation with SVM which is frequently used by data sets with missing values.

AB - The problem of missing values frequently occurs during data analysis. Imputation is one of the solutions to handle missing data. Clinical data often contain multiple measurements such as laboratory test results which are measured at different time points. In this study, we compared three imputation methods and their effects on different multiple measurement data sets with different sampling time periods. Data sets of liver cancer were used in this study for classification of liver cancer recurrence based on two types of classification models built by support vector machine (SVM) and random forests. The results report appropriate combinations of imputation methods and sampling time periods which achieve better classification results than those of other imputation methods and periods. These reported the leading imputation method with SVM is significantly different (P<0.001) from mean imputation with SVM which is frequently used by data sets with missing values.

KW - Missing values

KW - imputation

KW - random forests

KW - support vector machine

UR - http://www.scopus.com/inward/record.url?scp=84926444199&partnerID=8YFLogxK

U2 - 10.3233/978-1-61499-484-8-1930

DO - 10.3233/978-1-61499-484-8-1930

M3 - 会议稿件

AN - SCOPUS:84926444199

T3 - Frontiers in Artificial Intelligence and Applications

SP - 1930

EP - 1939

BT - Intelligent Systems and Applications - Proceedings of the International Computer Symposium, ICS 2014

A2 - Chu, William Cheng-Chung

A2 - Chao, Han-Chieh

A2 - Yang, Stephen Jenn-Hwa

PB - IOS Press BV

Y2 - 12 December 2014 through 14 December 2014

ER -

Ping XO, Tseng YJ, Liang JD, Huang GT, Yang PM, Lai F. Missing value imputation on multiple measurements for prediction of liver cancer recurrence: A comparative study. In Chu WCC, Chao HC, Yang SJH, editors, Intelligent Systems and Applications - Proceedings of the International Computer Symposium, ICS 2014. IOS Press BV. 2015. p. 1930-1939. (Frontiers in Artificial Intelligence and Applications). doi: 10.3233/978-1-61499-484-8-1930

Missing value imputation on multiple measurements for prediction of liver cancer recurrence: A comparative study

Abstract

Publication series

Conference

Bibliographical note

Keywords

UN SDGs

Access to Document

Other files and links

Fingerprint

Cite this