A peek into the black box: exploring classifiers by randomization. Henelius, A., Puolamäki, K., Boström, H., Asker, L., & Papapetrou, P. Data Min Knowl Disc, 28(5-6):1503--1529, September, 2014. ![A peek into the black box: exploring classifiers by randomization [link]](https://bibbase.org/img/filetypes/link.svg "link")
Paper doi abstract bibtex Classifiers are often opaque and cannot easily be inspected to gain understanding of which factors are of importance. We propose an efficient iterative algorithm to find the attributes and dependencies used by any classifier when making predictions. The performance and utility of the algorithm is demonstrated on two synthetic and 26 real-world datasets, using 15 commonly used learning algorithms to generate the classifiers. The empirical investigation shows that the novel algorithm is indeed able to find groupings of interacting attributes exploited by the different classifiers. These groupings allow for finding similarities among classifiers for a single dataset as well as for determining the extent to which different classifiers exploit such interactions in general.
@article{henelius_peek_2014,
title = {A peek into the black box: exploring classifiers by randomization},
volume = {28},
issn = {1384-5810, 1573-756X},
shorttitle = {A peek into the black box},
url = {http://link.springer.com/article/10.1007/s10618-014-0368-8},
doi = {10.1007/s10618-014-0368-8},
abstract = {Classifiers are often opaque and cannot easily be inspected to gain understanding of which factors are of importance. We propose an efficient iterative algorithm to find the attributes and dependencies used by any classifier when making predictions. The performance and utility of the algorithm is demonstrated on two synthetic and 26 real-world datasets, using 15 commonly used learning algorithms to generate the classifiers. The empirical investigation shows that the novel algorithm is indeed able to find groupings of interacting attributes exploited by the different classifiers. These groupings allow for finding similarities among classifiers for a single dataset as well as for determining the extent to which different classifiers exploit such interactions in general.},
language = {en},
number = {5-6},
urldate = {2016-12-14},
journal = {Data Min Knowl Disc},
author = {Henelius, Andreas and Puolamäki, Kai and Boström, Henrik and Asker, Lars and Papapetrou, Panagiotis},
month = sep,
year = {2014},
pages = {1503--1529},
file = {Full Text PDF:C\:\\Users\\Ashudeep Singh\\Zotero\\storage\\ET77A6A5\\Henelius et al. - 2014 - A peek into the black box exploring classifiers b.pdf:application/pdf;Snapshot:C\:\\Users\\Ashudeep Singh\\Zotero\\storage\\SPVUX9PC\\s10618-014-0368-8.html:text/html}
}
Downloads: 0
{"_id":"MkYmBMDhaLCjvmm9i","bibbaseid":"henelius-puolamki-bostrm-asker-papapetrou-apeekintotheblackboxexploringclassifiersbyrandomization-2014","downloads":0,"creationDate":"2018-01-23T22:21:13.720Z","title":"A peek into the black box: exploring classifiers by randomization","author_short":["Henelius, A.","Puolamäki, K.","Boström, H.","Asker, L.","Papapetrou, P."],"year":2014,"bibtype":"article","biburl":"http://www.ashudeepsingh.com/FATML-Bibliography/fatcs.bib","bibdata":{"bibtype":"article","type":"article","title":"A peek into the black box: exploring classifiers by randomization","volume":"28","issn":"1384-5810, 1573-756X","shorttitle":"A peek into the black box","url":"http://link.springer.com/article/10.1007/s10618-014-0368-8","doi":"10.1007/s10618-014-0368-8","abstract":"Classifiers are often opaque and cannot easily be inspected to gain understanding of which factors are of importance. We propose an efficient iterative algorithm to find the attributes and dependencies used by any classifier when making predictions. The performance and utility of the algorithm is demonstrated on two synthetic and 26 real-world datasets, using 15 commonly used learning algorithms to generate the classifiers. The empirical investigation shows that the novel algorithm is indeed able to find groupings of interacting attributes exploited by the different classifiers. These groupings allow for finding similarities among classifiers for a single dataset as well as for determining the extent to which different classifiers exploit such interactions in general.","language":"en","number":"5-6","urldate":"2016-12-14","journal":"Data Min Knowl Disc","author":[{"propositions":[],"lastnames":["Henelius"],"firstnames":["Andreas"],"suffixes":[]},{"propositions":[],"lastnames":["Puolamäki"],"firstnames":["Kai"],"suffixes":[]},{"propositions":[],"lastnames":["Boström"],"firstnames":["Henrik"],"suffixes":[]},{"propositions":[],"lastnames":["Asker"],"firstnames":["Lars"],"suffixes":[]},{"propositions":[],"lastnames":["Papapetrou"],"firstnames":["Panagiotis"],"suffixes":[]}],"month":"September","year":"2014","pages":"1503--1529","file":"Full Text PDF:C\\:\\\\Users\\\\Ashudeep Singh\\\\Zotero\\\\storage\\\\ET77A6A5\\e̋nelius et al. - 2014 - A peek into the black box exploring classifiers b.pdf:application/pdf;Snapshot:C\\:\\\\Users\\\\Ashudeep Singh\\\\Zotero\\\\storage\\\\SPVUX9PC\\\\s10618-014-0368-8.html:text/html","bibtex":"@article{henelius_peek_2014,\n\ttitle = {A peek into the black box: exploring classifiers by randomization},\n\tvolume = {28},\n\tissn = {1384-5810, 1573-756X},\n\tshorttitle = {A peek into the black box},\n\turl = {http://link.springer.com/article/10.1007/s10618-014-0368-8},\n\tdoi = {10.1007/s10618-014-0368-8},\n\tabstract = {Classifiers are often opaque and cannot easily be inspected to gain understanding of which factors are of importance. We propose an efficient iterative algorithm to find the attributes and dependencies used by any classifier when making predictions. The performance and utility of the algorithm is demonstrated on two synthetic and 26 real-world datasets, using 15 commonly used learning algorithms to generate the classifiers. The empirical investigation shows that the novel algorithm is indeed able to find groupings of interacting attributes exploited by the different classifiers. These groupings allow for finding similarities among classifiers for a single dataset as well as for determining the extent to which different classifiers exploit such interactions in general.},\n\tlanguage = {en},\n\tnumber = {5-6},\n\turldate = {2016-12-14},\n\tjournal = {Data Min Knowl Disc},\n\tauthor = {Henelius, Andreas and Puolamäki, Kai and Boström, Henrik and Asker, Lars and Papapetrou, Panagiotis},\n\tmonth = sep,\n\tyear = {2014},\n\tpages = {1503--1529},\n\tfile = {Full Text PDF:C\\:\\\\Users\\\\Ashudeep Singh\\\\Zotero\\\\storage\\\\ET77A6A5\\\\Henelius et al. - 2014 - A peek into the black box exploring classifiers b.pdf:application/pdf;Snapshot:C\\:\\\\Users\\\\Ashudeep Singh\\\\Zotero\\\\storage\\\\SPVUX9PC\\\\s10618-014-0368-8.html:text/html}\n}\n\n","author_short":["Henelius, A.","Puolamäki, K.","Boström, H.","Asker, L.","Papapetrou, P."],"key":"henelius_peek_2014","id":"henelius_peek_2014","bibbaseid":"henelius-puolamki-bostrm-asker-papapetrou-apeekintotheblackboxexploringclassifiersbyrandomization-2014","role":"author","urls":{"Paper":"http://link.springer.com/article/10.1007/s10618-014-0368-8"},"downloads":0,"html":""},"search_terms":["peek","black","box","exploring","classifiers","randomization","henelius","puolamäki","boström","asker","papapetrou"],"keywords":[],"authorIDs":[],"dataSources":["uzaKuqTDDSn4hnmBP"]}