Identification of signatures in biomedical spectra using domain knowledge

Identification of signatures in biomedical spectra using domain knowledge. Pranckeviciene, E., Somorjai, R., Baumgartner, R., & Jeon, M. G. Artificial Intelligence in Medicine, 35(3):215–226, Nov, 2005.

Link abstract bibtex

Demonstrate that incorporating domain knowledge into feature selection methods helps identify interpretable features with predictive capability comparable to a state-of-the-art classifier.\\ Two feature selection methods, one using a genetic algorithm (GA) the other a L(1)-norm support vector machine (SVM), were investigated on three real-world biomedical magnetic resonance (MR) spectral datasets of increasing difficulty. Consensus sets of the feature sets obtained by the two methods were also assessed.\\ Features identified independently by the two methods and by their consensus, determine class-discriminatory groups or individual features, whose predictive power compares favorably with that of a state-of-the-art classifier. Furthermore, the identified feature signatures form stable groupings at definite spectral positions, hence are readily interpretable. This is a useful and important practical result for generating hypothesis for the domain expert.

@Article{pmid16311187,
   Author="Pranckeviciene, E.  and Somorjai, R.  and Baumgartner, R.  and Jeon, M. G. ",
   Title="{{I}dentification of signatures in biomedical spectra using domain knowledge}",
   Journal="Artificial Intelligence in Medicine",
   Year="2005",
   Volume="35",
   Number="3",
   Pages="215--226",
   Month="Nov",
   Abstract={Demonstrate that incorporating domain knowledge into feature selection methods helps identify interpretable features with predictive capability comparable to a state-of-the-art classifier.\\ Two feature selection methods, one using a genetic algorithm (GA) the other a L(1)-norm support vector machine (SVM), were investigated on three real-world biomedical magnetic resonance (MR) spectral datasets of increasing difficulty. Consensus sets of the feature sets obtained by the two methods were also assessed.\\ Features identified independently by the two methods and by their consensus, determine class-discriminatory groups or individual features, whose predictive power compares favorably with that of a state-of-the-art classifier. Furthermore, the identified feature signatures form stable groupings at definite spectral positions, hence are readily interpretable. This is a useful and important practical result for generating hypothesis for the domain expert.},
   keywords={classification of biomedical spectra, dimensionality reduction, feature selection, genetic algorithm, L1-norm SVM, spectral signature, consensus feature sets, domain knowledge},
url_Link={http://www.aiimjournal.com/article/S0933-3657(05)00012-6/abstract?cc=y=}
%url_Paper={http://erinijapranckeviciene.net/biomedical_projects/Using_domain_knowledge_AImed2005.pdf}

}

Downloads: 0

{"_id":"Spo9q4dTQ8q6o5MLQ","bibbaseid":"pranckeviciene-somorjai-baumgartner-jeon-identificationofsignaturesinbiomedicalspectrausingdomainknowledge-2005","downloads":0,"creationDate":"2017-02-13T03:40:38.576Z","title":"Identification of signatures in biomedical spectra using domain knowledge","author_short":["Pranckeviciene, E.","Somorjai, R.","Baumgartner, R.","Jeon, M. G."],"year":2005,"bibtype":"article","biburl":"http://erinijapranckeviciene.net/erinija_publications.bib","bibdata":{"bibtype":"article","type":"article","author":[{"propositions":[],"lastnames":["Pranckeviciene"],"firstnames":["E."],"suffixes":[]},{"propositions":[],"lastnames":["Somorjai"],"firstnames":["R."],"suffixes":[]},{"propositions":[],"lastnames":["Baumgartner"],"firstnames":["R."],"suffixes":[]},{"propositions":[],"lastnames":["Jeon"],"firstnames":["M.","G."],"suffixes":[]}],"title":"Identification of signatures in biomedical spectra using domain knowledge","journal":"Artificial Intelligence in Medicine","year":"2005","volume":"35","number":"3","pages":"215–226","month":"Nov","abstract":"Demonstrate that incorporating domain knowledge into feature selection methods helps identify interpretable features with predictive capability comparable to a state-of-the-art classifier.\\\\ Two feature selection methods, one using a genetic algorithm (GA) the other a L(1)-norm support vector machine (SVM), were investigated on three real-world biomedical magnetic resonance (MR) spectral datasets of increasing difficulty. Consensus sets of the feature sets obtained by the two methods were also assessed.\\\\ Features identified independently by the two methods and by their consensus, determine class-discriminatory groups or individual features, whose predictive power compares favorably with that of a state-of-the-art classifier. Furthermore, the identified feature signatures form stable groupings at definite spectral positions, hence are readily interpretable. This is a useful and important practical result for generating hypothesis for the domain expert.","keywords":"classification of biomedical spectra, dimensionality reduction, feature selection, genetic algorithm, L1-norm SVM, spectral signature, consensus feature sets, domain knowledge","url_link":"http://www.aiimjournal.com/article/S0933-3657(05)00012-6/abstract?cc=y=","bibtex":"@Article{pmid16311187,\r\n Author=\"Pranckeviciene, E. and Somorjai, R. and Baumgartner, R. and Jeon, M. G. \",\r\n Title=\"{{I}dentification of signatures in biomedical spectra using domain knowledge}\",\r\n Journal=\"Artificial Intelligence in Medicine\",\r\n Year=\"2005\",\r\n Volume=\"35\",\r\n Number=\"3\",\r\n Pages=\"215--226\",\r\n Month=\"Nov\",\r\n Abstract={Demonstrate that incorporating domain knowledge into feature selection methods helps identify interpretable features with predictive capability comparable to a state-of-the-art classifier.\\\\ Two feature selection methods, one using a genetic algorithm (GA) the other a L(1)-norm support vector machine (SVM), were investigated on three real-world biomedical magnetic resonance (MR) spectral datasets of increasing difficulty. Consensus sets of the feature sets obtained by the two methods were also assessed.\\\\ Features identified independently by the two methods and by their consensus, determine class-discriminatory groups or individual features, whose predictive power compares favorably with that of a state-of-the-art classifier. Furthermore, the identified feature signatures form stable groupings at definite spectral positions, hence are readily interpretable. This is a useful and important practical result for generating hypothesis for the domain expert.},\r\n keywords={classification of biomedical spectra, dimensionality reduction, feature selection, genetic algorithm, L1-norm SVM, spectral signature, consensus feature sets, domain knowledge},\r\nurl_Link={http://www.aiimjournal.com/article/S0933-3657(05)00012-6/abstract?cc=y=}\r\n%url_Paper={http://erinijapranckeviciene.net/biomedical_projects/Using_domain_knowledge_AImed2005.pdf}\r\n\r\n}\r\n\r\n","author_short":["Pranckeviciene, E.","Somorjai, R.","Baumgartner, R.","Jeon, M. G."],"key":"pmid16311187","id":"pmid16311187","bibbaseid":"pranckeviciene-somorjai-baumgartner-jeon-identificationofsignaturesinbiomedicalspectrausingdomainknowledge-2005","role":"author","urls":{" link":"http://www.aiimjournal.com/article/S0933-3657(05)00012-6/abstract?cc=y="},"keyword":["classification of biomedical spectra","dimensionality reduction","feature selection","genetic algorithm","L1-norm SVM","spectral signature","consensus feature sets","domain knowledge"],"downloads":0,"html":""},"search_terms":["identification","signatures","biomedical","spectra","using","domain","knowledge","pranckeviciene","somorjai","baumgartner","jeon"],"keywords":["classification of biomedical spectra","dimensionality reduction","feature selection","genetic algorithm","l1-norm svm","spectral signature","consensus feature sets","domain knowledge"],"authorIDs":[],"dataSources":["wyhwdvMCTbALq9k63"]}