{"type":"ScholarlyArticle","title":"On Robust Trimming of Bayesian Network Classifiers","authors":[{"name":"YooJung Choi","slug":"yoojung-choi","given":"YooJung","family":"Choi","affiliation":"Computer Science Department , University of California , Los Angeles"},{"name":"Guy Van den Broeck","slug":"guy-van-den-broeck","given":"Guy","family":"Van den Broeck","affiliation":"Computer Science Department , University of California , Los Angeles"}],"abstract":"This paper considers the problem of removing costly features from a Bayesian network classifier. We want the classifier to be robust to these changes, and maintain its classification behavior. To this end, we propose a closeness metric between Bayesian classifiers, called the expected classification agreement (ECA). Our corresponding trimming algorithm finds an optimal subset of features and a new classification threshold that maximize the expected agreement, subject to a budgetary constraint. It utilizes new theoretical insights to perform branch-and-bound search in the space of feature sets, while computing bounds on the ECA. Our experiments investigate both the runtime cost of trimming and its effect on the robustness and accuracy of the final classifier.","doi":"10.24963/ijcai.2018/694","url":"https://ijcai.org/proceedings/2018/694","pdf":"https://ijcai.org/proceedings/2018/0694.pdf","markdown":"https://ijcai.org/proceedings/2018/694.md","bibtex":"https://ijcai.org/proceedings/2018/bibtex/694","volume":{"slug":"2018","title":"Proceedings of the Twenty-Seventh International Joint Conference on Artificial Intelligence","short_title":"IJCAI 2018","url":"https://ijcai.org/proceedings/2018/","year":2018,"location":"Stockholm","dates":"13-19 July 2018","editors":"Jérôme Lang","isbn":"978-0-9992411-2-7","publisher":"International Joint Conferences on Artificial Intelligence Organization","doi_prefix":"10.24963/ijcai.2018"},"section":"Uncertainty in AI","track":"Main track","number":694,"pages":{"first":5002,"last":5009},"keywords":["Heuristic Search and Game Playing: Combinatorial Search and Optimisation","Machine Learning: Feature Selection ; Learning Sparse Models","Uncertainty in AI: Bayesian Networks","Uncertainty in AI: Exact Probabilistic Inference"],"videos":[],"references":[{"position":1,"raw":"[Bache and Lichman, 2013] K. Bache and M. Lichman. UCI machine learning repository, 2013.","title":"UCI machine learning repository","authors":"K Bache, M Lichman","year":"2013","venue":null,"doi":null,"url":null,"internal_url":null},{"position":2,"raw":"[Bellazzi and Zupan, 2008] Riccardo Bellazzi and Blaz Zu- pan. Predictive data mining in clinical medicine: current issues and guidelines. International journal of medical informatics, 77(2):81-97, 2008.","title":"Riccardo Bellazzi and Blaz Zupan. Predictive data mining in clinical medicine: current issues and guidelines","authors":"Zupan Bellazzi","year":"2008","venue":"International journal of medical informatics","doi":null,"url":null,"internal_url":null},{"position":3,"raw":"[Bilgic and Getoor, 2011] Mustafa Bilgic and Lise Getoor. Value of information lattice: Exploiting probabilistic inde- pendence for effective feature subset acquisition. Journal of Artificial Intelligence Research (JAIR), 41:69-95, 2011.","title":"Value of information lattice: Exploiting probabilistic independence for effective feature subset acquisition","authors":"Mustafa Bilgic, Lise Getoor","year":"2011","venue":"Journal of Artificial Intelligence Research (JAIR)","doi":null,"url":null,"internal_url":null},{"position":4,"raw":"[Boutilier et al., 1996] Craig Boutilier, Nir Friedman, Moises Goldszmidt, and Daphne Koller. Context-specific indepen- dence in bayesian networks. In Proceedings of the Twelfth international conference on Uncertainty in artificial intelli- gence, pages 115-123.","title":"Context-specific independence in bayesian networks","authors":"Craig Boutilier, Nir Friedman, Moises Goldszmidt, Daphne Koller","year":null,"venue":"Proceedings of the Twelfth international conference on Uncertainty in artificial intelligence","doi":null,"url":null,"internal_url":null},{"position":5,"raw":"[Chavira and Darwiche, 2008] Mark Chavira and Adnan Dar- wiche. On probabilistic inference by weighted model count- ing. Artificial Intelligence, 172(6-7):772-799, 2008.","title":"Mark Chavira and Adnan Darwiche. On probabilistic inference by weighted model counting","authors":"Darwiche Chavira","year":"2008","venue":"Artificial Intelligence","doi":null,"url":null,"internal_url":null},{"position":6,"raw":"[Chen et al., 2015] Suming Chen, Arthur Choi, and Adnan Darwiche. Value of information based on Decision Robust- ness. In Proceedings of the 29th Conference on Artificial Intelligence (AAAI), 2015.","title":"Value of information based on Decision Robustness","authors":"Chen","year":"2015","venue":"Proceedings of the 29th Conference on Artificial Intelligence (AAAI)","doi":null,"url":null,"internal_url":null},{"position":7,"raw":"[Cheng and Greiner, 1999] Jie Cheng and Russell Greiner. Comparing bayesian network classifiers. In Proceedings of the Fifteenth conference on Uncertainty in artificial intelli- gence, pages 101-108.","title":"Comparing bayesian network classifiers","authors":"Greiner ; ] Jie Cheng, Russell Cheng, Greiner","year":"1999","venue":"Proceedings of the Fifteenth conference on Uncertainty in artificial intelligence","doi":null,"url":null,"internal_url":null},{"position":8,"raw":"[Choi et al., 2012] Arthur Choi, Yexiang Xue, and Adnan Darwiche. Same-Decision Probability: A confidence mea- sure for threshold-based decisions. International Journal of Approximate Reasoning (IJAR), 2:1415-1428, 2012.","title":"Same-Decision Probability: A confidence measure for threshold-based decisions","authors":"Arthur Choi, Yexiang Xue, Adnan Darwiche","year":"2012","venue":"International Journal of Approximate Reasoning (IJAR)","doi":null,"url":null,"internal_url":null},{"position":9,"raw":"[Choi et al., 2013] Arthur Choi, Doga Kisa, and Adnan Dar- wiche. Compiling probabilistic graphical models using sentential decision diagrams. In European Conference on Symbolic and Quantitative Approaches to Reasoning and Uncertainty, pages 121-132. Springer, 2013.","title":"Compiling probabilistic graphical models using sentential decision diagrams","authors":"Arthur Choi, Doga Kisa, Adnan Darwiche","year":"2013","venue":"European Conference on Symbolic and Quantitative Approaches to Reasoning and Uncertainty","doi":null,"url":null,"internal_url":null},{"position":10,"raw":"[Choi et al., 2017] YooJung Choi, Adnan Darwiche, and Guy Van den Broeck. Optimal feature selection for decision robustness in bayesian networks. In Proceedings of the 26th International Joint Conference on Artificial Intelligence (IJCAI), August 2017.","title":"Optimal feature selection for decision robustness in bayesian networks","authors":"Choi","year":"2017","venue":"Proceedings of the 26th International Joint Conference on Artificial Intelligence (IJCAI)","doi":null,"url":null,"internal_url":"/proceedings/2017/215"},{"position":11,"raw":"[Darwiche, 2011] Adnan Darwiche. SDD: A new canonical representation of propositional knowledge bases. In Pro- ceedings of the 22nd International Joint Conference on Artificial Intelligence, pages 819-826, 2011.","title":"SDD: A new canonical representation of propositional knowledge bases","authors":"Adnan Darwiche","year":"2011","venue":"Proceedings of the 22nd International Joint Conference on Artificial Intelligence","doi":null,"url":null,"internal_url":"/proceedings/2011/143"},{"position":12,"raw":"[Friedman et al., 1997] Nir Friedman, Dan Geiger, and Moi- ses Goldszmidt. Bayesian network classifiers. Machine learning, 29(2-3):131-163, 1997.","title":"Bayesian network classifiers","authors":"Nir Friedman, Dan Geiger, Moises Goldszmidt","year":"1997","venue":"Machine learning","doi":null,"url":null,"internal_url":null},{"position":13,"raw":"[Gao and Koller, 2011] Tianshi Gao and Daphne Koller. Ac- tive classification based on value of classifier. In Advances in Neural Information Processing Systems, pages 1062- 1070, 2011.","title":"Active classification based on value of classifier","authors":"Koller Gao, Tianshi Gao, Daphne Koller","year":"2011","venue":"Advances in Neural Information Processing Systems","doi":null,"url":null,"internal_url":null},{"position":14,"raw":"[Gimenez et al., 2014] Francisco J Gimenez, Yirong Wu, Elizabeth S Burnside, and Daniel L Rubin. A novel method to assess incompleteness of mammography reports. In AMIA Annual Symposium Proceedings, volume 2014, page 1758. American Medical Informatics Association, 2014.","title":"A novel method to assess incompleteness of mammography reports","authors":"Gimenez","year":"2014","venue":"AMIA Annual Symposium Proceedings","doi":null,"url":null,"internal_url":null},{"position":15,"raw":"[Kira and Rendell, 1992] Kenji Kira and Larry A Rendell. The feature selection problem: Traditional methods and a new algorithm. In AAAI, volume 2, pages 129-134, 1992.","title":"The feature selection problem: Traditional methods and a new algorithm","authors":"Kenji Kira, Larry A Rendell","year":"1992","venue":"AAAI","doi":null,"url":null,"internal_url":null},{"position":16,"raw":"[Kolesar, 1967] Peter J Kolesar. A branch and bound algo- rithm for the knapsack problem. Management science, 13(9):723-735, 1967.","title":"A branch and bound algorithm for the knapsack problem","authors":"J Peter, Kolesar","year":"1967","venue":"Management science","doi":null,"url":null,"internal_url":null},{"position":17,"raw":"[Kollar and Roy, 2008] Thomas Kollar and Nicholas Roy. Ef- ficient optimization of information-theoretic exploration in slam. In AAAI, volume 8, pages 1369-1375, 2008.","title":"Efficient optimization of information-theoretic exploration in slam","authors":"Roy ; Kollar, Thomas Kollar, Nicholas Roy","year":"2008","venue":"AAAI","doi":null,"url":null,"internal_url":null},{"position":18,"raw":"[Kononenko, 2001] Igor Kononenko. Machine learning for medical diagnosis: history, state of the art and perspective. Artificial Intelligence in medicine, 23(1):89-109, 2001.","title":"Machine learning for medical diagnosis: history, state of the art and perspective","authors":"Kononenko","year":"2001","venue":"Artificial Intelligence in medicine","doi":null,"url":null,"internal_url":null},{"position":19,"raw":"[Krause and Guestrin, 2009] Andreas Krause and Carlos Guestrin. Optimal value of information in graphical models. Journal of Artificial Intelligence Research (JAIR), 35:557- 591, 2009.","title":"Optimal value of information in graphical models","authors":"Guestrin Krause, Andreas Krause, Carlos Guestrin","year":"2009","venue":"Journal of Artificial Intelligence Research (JAIR)","doi":null,"url":null,"internal_url":null},{"position":20,"raw":"[Laidlaw et al., 1998] David H Laidlaw, Kurt W Fleischer, and Alan H Barr. Partial-volume bayesian classification of material mixtures in mr volume data using voxel histograms. IEEE transactions on medical imaging, 17(1):74-86, 1998.","title":"Partial-volume bayesian classification of material mixtures in mr volume data using voxel histograms","authors":"Kurt W David H Laidlaw, Alan H Fleischer, Barr","year":"1998","venue":"IEEE transactions on medical imaging","doi":null,"url":null,"internal_url":null},{"position":21,"raw":"[Lucas, 2001] Peter Lucas. Expert knowledge and its role in learning bayesian networks in medicine: an appraisal. In Conference on Artificial Intelligence in Medicine in Europe, pages 156-166. Springer, 2001.","title":"Expert knowledge and its role in learning bayesian networks in medicine: an appraisal","authors":"Lucas ;, Peter Lucas","year":"2001","venue":"Conference on Artificial Intelligence in Medicine in Europe","doi":null,"url":null,"internal_url":null},{"position":22,"raw":"[Metsis et al., 2006] Vangelis Metsis, Ion Androutsopoulos, and Georgios Paliouras. Spam filtering with naive bayes- which naive bayes? In CEAS, volume 17, pages 28-69, 2006.","title":"Vangelis Metsis, Ion Androutsopoulos, and Georgios Paliouras. Spam filtering with naive bayeswhich naive bayes?","authors":"Metsis","year":"2006","venue":"CEAS","doi":null,"url":null,"internal_url":null},{"position":23,"raw":"[Millán and Pérez-De-La-Cruz, 2002] Eva Millán and José Luis Pérez-De-La-Cruz. A Bayesian diagnostic algorithm for student modeling and its evaluation. User Modeling and User-Adapted Interaction, 12(2-3):281-330, 2002.","title":"Eva Millán and José Luis Pérez-De-La-Cruz. A Bayesian diagnostic algorithm for student modeling and its evaluation","authors":"Pérez-De-La-Cruz Millán","year":"2002","venue":"User Modeling and User-Adapted Interaction","doi":null,"url":null,"internal_url":null},{"position":24,"raw":"[Munie and Shoham, 2008] Michael Munie and Yoav Shoham. Optimal testing of structured knowledge. In Proceedings of the 23rd National Conference on Artificial intelligence, pages 1069-1074, 2008.","title":"Optimal testing of structured knowledge","authors":"Shoham Munie, Michael Munie, Yoav Shoham","year":"2008","venue":"Proceedings of the 23rd National Conference on Artificial intelligence","doi":null,"url":null,"internal_url":null},{"position":25,"raw":"[Narendra and Fukunaga, 1977] Patrenahalli M. Narendra and Keinosuke Fukunaga. A branch and bound algorithm for feature subset selection. IEEE Transactions on Comput- ers, 26(9):917-922, 1977.","title":"A branch and bound algorithm for feature subset selection","authors":"M Patrenahalli, Keinosuke Narendra, Fukunaga","year":"1977","venue":"IEEE Transactions on Computers","doi":null,"url":null,"internal_url":null},{"position":26,"raw":"[Wu et al., 2008] Xindong Wu, Vipin Kumar, J Ross Quinlan, Joydeep Ghosh, Qiang Yang, Hiroshi Motoda, Geoffrey J McLachlan, Angus Ng, Bing Liu, S Yu Philip, et al. Top 10 algorithms in data mining. Knowledge and information systems, 14(1):1-37, 2008.","title":"Top 10 algorithms in data mining","authors":"Xindong Wu, Vipin Kumar, J Ross Quinlan, Joydeep Ghosh, Qiang Yang, Hiroshi Motoda, Geoffrey J Mclachlan, Angus Ng, Bing Liu, S Yu Philip","year":"2008","venue":"Knowledge and information systems","doi":null,"url":null,"internal_url":null},{"position":27,"raw":"[Yu and Liu, 2004] Lei Yu and Huan Liu. Efficient feature selection via analysis of relevance and redundancy. Journal of machine learning research, 5(Oct):1205-1224, 2004.","title":"Efficient feature selection via analysis of relevance and redundancy","authors":"Yu, Liu ; ] Lei Yu, Huan Liu","year":"2004","venue":"Journal of machine learning research","doi":null,"url":null,"internal_url":null},{"position":28,"raw":"[Zhang and Ji, 2010] Yongmian Zhang and Qiang Ji. Effi- cient sensor selection for active information fusion. IEEE Transactions on Systems, Man, and Cybernetics, Part B, 40(3):719-728, June 2010.","title":"Yongmian Zhang and Qiang Ji. Efficient sensor selection for active information fusion","authors":"Ji Zhang","year":"2010","venue":"IEEE Transactions on Systems, Man, and Cybernetics, Part B","doi":null,"url":null,"internal_url":null},{"position":29,"raw":"[Zhao et al., 2017] Jieyu Zhao, Tianlu Wang, Mark Yatskar, Vicente Ordonez, and Kai-Wei Chang. Men also like shop- ping: Reducing gender bias amplification using corpus- level constraints. In EMNLP, 2017.","title":"Men also like shopping: Reducing gender bias amplification using corpuslevel constraints","authors":"Zhao","year":"2017","venue":"EMNLP","doi":null,"url":null,"internal_url":null}],"license_note":"Copyright International Joint Conferences on Artificial Intelligence."}