On the use of multiobjective evolutionary algorithms for survival analysis. 

Abstract  This paper proposes and evaluates a multiobjective evolutionary algorithm for survival analysis. One aim of survival analysis is the extraction of models from data that approximate lifetime/failure time distributions. These models can be used to estimate the time that an event takes to happen to an object. To use of multiobjective evolutionary algorithms for survival analysis has several advantages. They can cope with feature interactions, noisy data, and are capable of optimising several objectives. This is important, as model extraction is a multiobjective problem. It has at least two objectives, which are the extraction of accurate and simple models. Accurate models are required to achieve good predictions. Simple models are important to prevent overfitting, improve the transparency of the models, and to save computational resources. Although there is a plethora of evolutionary approaches to extract models for classification and regression, the presented approach is one of the first applied to survival analysis. The approach is evaluated on several artificial datasets and one medical dataset. It is shown that the approach is capable of producing accurate models, even for problems that violate some of the assumptions made by classical approaches. 
PMID  16762491 
Related Publications 
Exploring the runtime of an evolutionary algorithm for the multiobjective shortest path problem. Biclustering of microarray data with MOSPO based on crowding distance. Incorporating evolution of transcription factor binding sites into annotated alignments. Ideal evaluation from coevolution. 
Authors  
Mayor MeshTerms  
Keywords  
Journal Title  bio systems 
Publication Year Start  20070101 
%A Setzkorn, Christian; Taktak, Azzam F. G.; Damato, Bertil E. %T On the use of multiobjective evolutionary algorithms for survival analysis. %J Bio Systems, vol. 87, no. 1, pp. 3148 %D 01/2007 %V 87 %N 1 %M eng %B This paper proposes and evaluates a multiobjective evolutionary algorithm for survival analysis. One aim of survival analysis is the extraction of models from data that approximate lifetime/failure time distributions. These models can be used to estimate the time that an event takes to happen to an object. To use of multiobjective evolutionary algorithms for survival analysis has several advantages. They can cope with feature interactions, noisy data, and are capable of optimising several objectives. This is important, as model extraction is a multiobjective problem. It has at least two objectives, which are the extraction of accurate and simple models. Accurate models are required to achieve good predictions. Simple models are important to prevent overfitting, improve the transparency of the models, and to save computational resources. Although there is a plethora of evolutionary approaches to extract models for classification and regression, the presented approach is one of the first applied to survival analysis. The approach is evaluated on several artificial datasets and one medical dataset. It is shown that the approach is capable of producing accurate models, even for problems that violate some of the assumptions made by classical approaches. %K Algorithms, Evaluation Studies as Topic, Models, Molecular, Survival Analysis %P 31 %L 48 %Y 10.1016/j.biosystems.2006.03.002 %W PHY %G AUTHOR %R 2007.......87...31S
@Article{Setzkorn2007, author="Setzkorn, Christian and Taktak, Azzam F. G. and Damato, Bertil E.", title="On the use of multiobjective evolutionary algorithms for survival analysis.", journal="Bio Systems", year="2007", month="Jan", day="13", volume="87", number="1", pages="3148", keywords="Algorithms", keywords="Evaluation Studies as Topic", keywords="Models, Molecular", keywords="Survival Analysis", abstract="This paper proposes and evaluates a multiobjective evolutionary algorithm for survival analysis. One aim of survival analysis is the extraction of models from data that approximate lifetime/failure time distributions. These models can be used to estimate the time that an event takes to happen to an object. To use of multiobjective evolutionary algorithms for survival analysis has several advantages. They can cope with feature interactions, noisy data, and are capable of optimising several objectives. This is important, as model extraction is a multiobjective problem. It has at least two objectives, which are the extraction of accurate and simple models. Accurate models are required to achieve good predictions. Simple models are important to prevent overfitting, improve the transparency of the models, and to save computational resources. Although there is a plethora of evolutionary approaches to extract models for classification and regression, the presented approach is one of the first applied to survival analysis. The approach is evaluated on several artificial datasets and one medical dataset. It is shown that the approach is capable of producing accurate models, even for problems that violate some of the assumptions made by classical approaches.", issn="03032647", doi="10.1016/j.biosystems.2006.03.002", url="http://www.ncbi.nlm.nih.gov/pubmed/16762491", language="eng" }
%0 Journal Article %T On the use of multiobjective evolutionary algorithms for survival analysis. %A Setzkorn, Christian %A Taktak, Azzam F. G. %A Damato, Bertil E. %J Bio Systems %D 2007 %8 Jan 13 %V 87 %N 1 %@ 03032647 %G eng %F Setzkorn2007 %X This paper proposes and evaluates a multiobjective evolutionary algorithm for survival analysis. One aim of survival analysis is the extraction of models from data that approximate lifetime/failure time distributions. These models can be used to estimate the time that an event takes to happen to an object. To use of multiobjective evolutionary algorithms for survival analysis has several advantages. They can cope with feature interactions, noisy data, and are capable of optimising several objectives. This is important, as model extraction is a multiobjective problem. It has at least two objectives, which are the extraction of accurate and simple models. Accurate models are required to achieve good predictions. Simple models are important to prevent overfitting, improve the transparency of the models, and to save computational resources. Although there is a plethora of evolutionary approaches to extract models for classification and regression, the presented approach is one of the first applied to survival analysis. The approach is evaluated on several artificial datasets and one medical dataset. It is shown that the approach is capable of producing accurate models, even for problems that violate some of the assumptions made by classical approaches. %K Algorithms %K Evaluation Studies as Topic %K Models, Molecular %K Survival Analysis %U http://dx.doi.org/10.1016/j.biosystems.2006.03.002 %U http://www.ncbi.nlm.nih.gov/pubmed/16762491 %P 3148
PT Journal AU Setzkorn, C Taktak, AFG Damato, BE TI On the use of multiobjective evolutionary algorithms for survival analysis. SO Bio Systems JI BioSystems PD Jan PY 2007 BP 31 EP 48 VL 87 IS 1 DI 10.1016/j.biosystems.2006.03.002 LA eng DE Algorithms; Evaluation Studies as Topic; Models, Molecular; Survival Analysis AB This paper proposes and evaluates a multiobjective evolutionary algorithm for survival analysis. One aim of survival analysis is the extraction of models from data that approximate lifetime/failure time distributions. These models can be used to estimate the time that an event takes to happen to an object. To use of multiobjective evolutionary algorithms for survival analysis has several advantages. They can cope with feature interactions, noisy data, and are capable of optimising several objectives. This is important, as model extraction is a multiobjective problem. It has at least two objectives, which are the extraction of accurate and simple models. Accurate models are required to achieve good predictions. Simple models are important to prevent overfitting, improve the transparency of the models, and to save computational resources. Although there is a plethora of evolutionary approaches to extract models for classification and regression, the presented approach is one of the first applied to survival analysis. The approach is evaluated on several artificial datasets and one medical dataset. It is shown that the approach is capable of producing accurate models, even for problems that violate some of the assumptions made by classical approaches. ER
PMID 16762491 OWN  NLM STAT MEDLINE DA  20061218 DCOM 20070222 LR  20071115 IS  03032647 (Print) IS  03032647 (Linking) VI  87 IP  1 DP  2007 Jan TI  On the use of multiobjective evolutionary algorithms for survival analysis. PG  3148 AB  This paper proposes and evaluates a multiobjective evolutionary algorithm for survival analysis. One aim of survival analysis is the extraction of models from data that approximate lifetime/failure time distributions. These models can be used to estimate the time that an event takes to happen to an object. To use of multiobjective evolutionary algorithms for survival analysis has several advantages. They can cope with feature interactions, noisy data, and are capable of optimising several objectives. This is important, as model extraction is a multiobjective problem. It has at least two objectives, which are the extraction of accurate and simple models. Accurate models are required to achieve good predictions. Simple models are important to prevent overfitting, improve the transparency of the models, and to save computational resources. Although there is a plethora of evolutionary approaches to extract models for classification and regression, the presented approach is one of the first applied to survival analysis. The approach is evaluated on several artificial datasets and one medical dataset. It is shown that the approach is capable of producing accurate models, even for problems that violate some of the assumptions made by classical approaches. FAU  Setzkorn, Christian AU  Setzkorn C AD  Royal Liverpool University Hospital, Liverpool, United Kingdom. [email protected] FAU  Taktak, Azzam F G AU  Taktak AF FAU  Damato, Bertil E AU  Damato BE LA  eng PT  Journal Article DEP  20060313 PL  Ireland TA  Biosystems JT  Bio Systems JID  0430773 SB  IM MH  *Algorithms MH  Evaluation Studies as Topic MH  Models, Molecular MH  *Survival Analysis EDAT 2006/06/10 09:00 MHDA 2007/02/23 09:00 CRDT 2006/06/10 09:00 PHST 2005/12/07 [received] PHST 2006/03/03 [revised] PHST 2006/03/03 [accepted] PHST 2006/03/13 [aheadofprint] AID  S03032647(06)000463 [pii] AID  10.1016/j.biosystems.2006.03.002 [doi] PST  ppublish SO  Biosystems. 2007 Jan;87(1):3148. Epub 2006 Mar 13.
TY  JOUR AU  Setzkorn, Christian AU  Taktak, Azzam F. G. AU  Damato, Bertil E. PY  2007/Jan/13 TI  On the use of multiobjective evolutionary algorithms for survival analysis. T2  BioSystems JO  Bio Systems SP  31 EP  48 VL  87 IS  1 KW  Algorithms KW  Evaluation Studies as Topic KW  Models, Molecular KW  Survival Analysis N2  This paper proposes and evaluates a multiobjective evolutionary algorithm for survival analysis. One aim of survival analysis is the extraction of models from data that approximate lifetime/failure time distributions. These models can be used to estimate the time that an event takes to happen to an object. To use of multiobjective evolutionary algorithms for survival analysis has several advantages. They can cope with feature interactions, noisy data, and are capable of optimising several objectives. This is important, as model extraction is a multiobjective problem. It has at least two objectives, which are the extraction of accurate and simple models. Accurate models are required to achieve good predictions. Simple models are important to prevent overfitting, improve the transparency of the models, and to save computational resources. Although there is a plethora of evolutionary approaches to extract models for classification and regression, the presented approach is one of the first applied to survival analysis. The approach is evaluated on several artificial datasets and one medical dataset. It is shown that the approach is capable of producing accurate models, even for problems that violate some of the assumptions made by classical approaches. SN  03032647 UR  http://dx.doi.org/10.1016/j.biosystems.2006.03.002 UR  http://www.ncbi.nlm.nih.gov/pubmed/16762491 ID  Setzkorn2007 ER 
<?xml version="1.0" encoding="UTF8"?> <b:Sources SelectedStyle="" xmlns:b="http://schemas.openxmlformats.org/officeDocument/2006/bibliography" xmlns="http://schemas.openxmlformats.org/officeDocument/2006/bibliography" > <b:Source> <b:Tag>Setzkorn2007</b:Tag> <b:SourceType>ArticleInAPeriodical</b:SourceType> <b:Year>2007</b:Year> <b:Month>Jan</b:Month> <b:Day>13</b:Day> <b:PeriodicalName>Bio Systems</b:PeriodicalName> <b:Volume>87</b:Volume> <b:Issue>1</b:Issue> <b:Pages>3148</b:Pages> <b:Author> <b:Author><b:NameList> <b:Person><b:Last>Setzkorn</b:Last><b:First>Christian</b:First></b:Person> <b:Person><b:Last>Taktak</b:Last><b:First>Azzam</b:First><b:Middle>F</b:Middle><b:Middle>G</b:Middle></b:Person> <b:Person><b:Last>Damato</b:Last><b:First>Bertil</b:First><b:Middle>E</b:Middle></b:Person> </b:NameList></b:Author> </b:Author> <b:Title>On the use of multiobjective evolutionary algorithms for survival analysis.</b:Title> <b:ShortTitle>BioSystems</b:ShortTitle> <b:Comments>This paper proposes and evaluates a multiobjective evolutionary algorithm for survival analysis. One aim of survival analysis is the extraction of models from data that approximate lifetime/failure time distributions. These models can be used to estimate the time that an event takes to happen to an object. To use of multiobjective evolutionary algorithms for survival analysis has several advantages. They can cope with feature interactions, noisy data, and are capable of optimising several objectives. This is important, as model extraction is a multiobjective problem. It has at least two objectives, which are the extraction of accurate and simple models. Accurate models are required to achieve good predictions. Simple models are important to prevent overfitting, improve the transparency of the models, and to save computational resources. Although there is a plethora of evolutionary approaches to extract models for classification and regression, the presented approach is one of the first applied to survival analysis. The approach is evaluated on several artificial datasets and one medical dataset. It is shown that the approach is capable of producing accurate models, even for problems that violate some of the assumptions made by classical approaches.</b:Comments> </b:Source> </b:Sources>