{"type":"ScholarlyArticle","title":"Learning Latest Classifiers without Additional Labeled Data","authors":[{"name":"Atsutoshi Kumagai","slug":"atsutoshi-kumagai","given":"Atsutoshi","family":"Kumagai","affiliation":"NTT Secure Platform Laboratories"},{"name":"Tomoharu Iwata","slug":"tomoharu-iwata","given":"Tomoharu","family":"Iwata","affiliation":"NTT Communication Science Laboratories"}],"abstract":"In various applications such as spam mail classification, the performance of classifiers deteriorates over time. Although retraining classifiers using labeled data helps to maintain the performance, continuously preparing labeled data is quite expensive. In this paper, we propose a method to learn classifiers by using newly obtained unlabeled data, which are easy to prepare, as well as labeled data collected beforehand. A major reason for the performance deterioration is the emergence of new features that do not appear in the training phase. Another major reason is the change of the distribution between the training and test phases. The proposed method learns the latest classifiers that overcome both problems. With the proposed method, the conditional distribution of new features given existing features is learned using the unlabeled data. In addition, the proposed method estimates the density ratio between training and test distributions by using the labeled and unlabeled data. We approximate the classification error of a classifier, which exploits new features as well as existing features, at the test phase by incorporating both the conditional distribution of new features and the densityratio, simultaneously. By minimizing the approximated error while integrating out new feature values, we obtain a classifier that exploits new features and fits on the test phase. The effectiveness of the proposed method is demonstrated with experiments using synthetic and real-world data sets.","doi":"10.24963/ijcai.2017/283","url":"https://ijcai.org/proceedings/2017/283","pdf":"https://ijcai.org/proceedings/2017/0283.pdf","markdown":"https://ijcai.org/proceedings/2017/283.md","bibtex":"https://ijcai.org/proceedings/2017/bibtex/283","volume":{"slug":"2017","title":"Proceedings of the Twenty-Sixth International Joint Conference on Artificial Intelligence","short_title":"IJCAI 2017","url":"https://ijcai.org/proceedings/2017/","year":2017,"location":"Melbourne, Australia","dates":"19-25 August 2017","editors":"Carles Sierra","isbn":"978-0-9992411-0-3","publisher":"International Joint Conferences on Artificial Intelligence Organization","doi_prefix":"10.24963/ijcai.2017"},"section":"Machine Learning A-R","track":"Main track","number":283,"pages":{"first":2039,"last":2045},"keywords":["Machine Learning: Classification","Machine Learning: Transfer, Adaptation, Multi-task Learning","Machine Learning: Machine Learning"],"videos":[],"references":[{"position":1,"raw":"[Abdallah et al., 2012] Zahraa Said Abdallah, Mo- hamed Medhat Gaber, Bama Srinivasan, and Shonali Krishnaswamy. Streamar: incremental and active learning with evolving sensory data for activity recognition. In ICTAI, 2012.","title":"Streamar: incremental and active learning with evolving sensory data for activity recognition","authors":"Zahraa Said Abdallah, Mohamed Medhat Gaber, Bama Srinivasan, Shonali Krishnaswamy","year":"2012","venue":"ICTAI","doi":null,"url":null,"internal_url":null},{"position":2,"raw":"[Bickel and Scheffer, 2007] Steffen Bickel and Tobias Scheffer. Dirichlet-enhanced spam filtering based on biased samples. NIPS, 2007.","title":"Dirichlet-enhanced spam filtering based on biased samples","authors":"Scheffer Bickel, Steffen Bickel, Tobias Scheffer","year":"2007","venue":"NIPS","doi":null,"url":null,"internal_url":null},{"position":3,"raw":"[Bickel et al., 2008] Steffen Bickel, Jasmina Bogojeska, Thomas Lengauer, and Tobias Scheffer. Multi-task learn- ing for hiv therapy screening. In ICML, 2008.","title":"Multi-task learning for hiv therapy screening","authors":"Steffen Bickel, Jasmina Bogojeska, Thomas Lengauer, Tobias Scheffer","year":"2008","venue":"ICML","doi":null,"url":null,"internal_url":null},{"position":4,"raw":"[Bishop, 2006] Christopher M Bishop. Pattern recognition and machine learning. springer, 2006.","title":"Pattern recognition and machine learning","authors":"M Christopher, Bishop","year":"2006","venue":null,"doi":null,"url":null,"internal_url":null},{"position":5,"raw":"[Crammer et al., 2009] Koby Crammer, Alex Kulesza, and Mark Dredze. Adaptive regularization of weight vectors. In NIPS, 2009.","title":"Adaptive regularization of weight vectors","authors":"Koby Crammer, Alex Kulesza, Mark Dredze","year":"2009","venue":"NIPS","doi":null,"url":null,"internal_url":null},{"position":6,"raw":"[Donders et al., 2006] A Rogier T Donders, Geert JMG van der Heijden, Theo Stijnen, and Karel GM Moons. Re- view: a gentle introduction to imputation of missing val- ues. JCE, 2006.","title":"Review: a gentle introduction to imputation of missing values","authors":"T Rogier, Donders, Jmg Geert, Theo Van Der Heijden, Stijnen, G M Karel, Moons","year":"2006","venue":null,"doi":null,"url":null,"internal_url":null},{"position":7,"raw":"[Fdez-Riverola et al., 2007] Florentino Fdez-Riverola, Eva Lorenzo Iglesias, Fernando Díaz, José Ramon Méndez, and Juan M Corchado. Applying lazy learning algorithms to tackle concept drift in spam filtering. ESWA, 2007.","title":"Applying lazy learning algorithms to tackle concept drift in spam filtering","authors":"Florentino Fdez-Riverola, Eva Lorenzo Iglesias, Fernando Díaz, José Ramon Méndez, Juan M Corchado","year":"2007","venue":null,"doi":null,"url":null,"internal_url":null},{"position":8,"raw":"[Gama et al., 2014] João Gama, Indrė Žliobaitė, Albert Bifet, Mykola Pechenizkiy, and Abdelhamid Bouchachia. A survey on concept drift adaptation. CSUR, 2014.","title":"A survey on concept drift adaptation","authors":"João Gama, Indrė Žliobaitė, Albert Bifet, Mykola Pechenizkiy, Abdelhamid Bouchachia","year":"2014","venue":"CSUR","doi":null,"url":null,"internal_url":null},{"position":9,"raw":"[Grandvalet and Bengio, 2004] Yves Grandvalet and Yoshua Bengio. Semi-supervised learning by entropy minimiza- tion. In NIPS, 2004.","title":"Semi-supervised learning by entropy minimization","authors":"Bengio Grandvalet, Yves Grandvalet, Yoshua Bengio","year":"2004","venue":"NIPS","doi":null,"url":null,"internal_url":null},{"position":10,"raw":"[Hachiya et al., 2012] Hirotaka Hachiya, Masashi Sugiyama, and Naonori Ueda. Importance-weighted least-squares probabilistic classifier for covariate shift adaptation with application to human activity recognition. Neurocomputing, 2012.","title":"Importance-weighted least-squares probabilistic classifier for covariate shift adaptation with application to human activity recognition","authors":"Hirotaka Hachiya, Masashi Sugiyama, Naonori Ueda","year":"2012","venue":"Neurocomputing","doi":null,"url":null,"internal_url":null},{"position":11,"raw":"[Haque et al., 2016] Ahsanul Haque, Latifur Khan, and Michael Baron. Sand: Semi-supervised adaptive novel class detection and classification over data stream. In AAAI, 2016.","title":"Sand: Semi-supervised adaptive novel class detection and classification over data stream","authors":"Haque","year":"2016","venue":"AAAI","doi":null,"url":null,"internal_url":null},{"position":12,"raw":"[Huang et al., 2006] Jiayuan Huang, Arthur Gretton, Karsten M Borgwardt, Bernhard Schölkopf, and Alex J Smola. Correcting sample selection bias by unlabeled data. In NIPS, 2006.","title":"Correcting sample selection bias by unlabeled data","authors":"Jiayuan Huang, Arthur Gretton, Karsten M Borgwardt, Bernhard Schölkopf, Alex J Smola","year":"2006","venue":"NIPS","doi":null,"url":null,"internal_url":null},{"position":13,"raw":"[Kanamori et al., 2009] Takafumi Kanamori, Shohei Hido, and Masahi Sugiyama. Efficient direct density ratio es- timation for non-stationarity adaptation and outlier detec- tion. In NIPS, 2009.","title":"Efficient direct density ratio estimation for non-stationarity adaptation and outlier detection","authors":"Takafumi Kanamori, Shohei Hido, Masahi Sugiyama","year":"2009","venue":"NIPS","doi":null,"url":null,"internal_url":null},{"position":14,"raw":"[Kingma et al., 2014] Diederik P Kingma, Shakir Mohamed, Danilo Jimenez Rezende, and Max Welling. Semi- supervised learning with deep generative models. In NIPS, 2014.","title":"Semisupervised learning with deep generative models","authors":"Shakir Diederik P Kingma, Danilo Mohamed, Max Jimenez Rezende, Welling","year":"2014","venue":"NIPS","doi":null,"url":null,"internal_url":null},{"position":15,"raw":"[Klinkenberg, 2004] Ralf Klinkenberg. Learning drifting concepts: Example selection vs. example weighting. In- telligent Data Analysis, 2004.","title":"Learning drifting concepts: Example selection vs. example weighting","authors":"Ralf Klinkenberg, Klinkenberg","year":"2004","venue":"Intelligent Data Analysis","doi":null,"url":null,"internal_url":null},{"position":16,"raw":"[Kumagai and Iwata, 2016] Atsutoshi Kumagai and Tomo- haru Iwata. Learning future classifiers without additional data. In AAAI, 2016.","title":"Atsutoshi Kumagai and Tomoharu Iwata. Learning future classifiers without additional data","authors":"Iwata Kumagai","year":"2016","venue":null,"doi":null,"url":null,"internal_url":null},{"position":17,"raw":"[Kumagai and Iwata, 2017] Atsutoshi Kumagai and Tomo- haru Iwata. Learning non-linear dynamics of decision boundaries for maintaining classification performance. In AAAI, 2017.","title":"Atsutoshi Kumagai and Tomoharu Iwata. Learning non-linear dynamics of decision boundaries for maintaining classification performance","authors":"Iwata Kumagai","year":"2017","venue":"AAAI","doi":null,"url":null,"internal_url":null},{"position":18,"raw":"[Li et al., 2010] Yan Li, Hiroyuki Kambara, Yasuharu Koike, and Masashi Sugiyama. Application of covariate shift adaptation techniques in brain-computer interfaces. Biomedical Engineering, IEEE Transactions on, 2010.","title":"Application of covariate shift adaptation techniques in brain-computer interfaces","authors":"Yan Li, Hiroyuki Kambara, Yasuharu Koike, Masashi Sugiyama","year":"2010","venue":"Biomedical Engineering, IEEE Transactions on","doi":null,"url":null,"internal_url":null},{"position":19,"raw":"[Liu and Ziebart, 2014] Anqi Liu and Brian Ziebart. Robust classification under sample selection bias. In NIPS. 2014.","title":"Robust classification under sample selection bias","authors":"Ziebart Liu, Anqi Liu, Brian Ziebart","year":"2014","venue":"NIPS","doi":null,"url":null,"internal_url":null},{"position":20,"raw":"[Ma et al., 2009] Justin Ma, Lawrence K Saul, Stefan Sav- age, and Geoffrey M Voelker. Identifying suspicious urls: an application of large-scale online learning. In ICML, 2009.","title":"Identifying suspicious urls: an application of large-scale online learning","authors":"Justin Ma, Lawrence K Saul, Stefan Savage, Geoffrey M Voelker","year":"2009","venue":"ICML","doi":null,"url":null,"internal_url":null},{"position":21,"raw":"[Nigam et al., 2000] Kamal Nigam, Andrew Kachites Mc- Callum, Sebastian Thrun, and Tom Mitchell. Text classi- fication from labeled and unlabeled documents using em. Machine Learning, 2000.","title":"Text classification from labeled and unlabeled documents using em","authors":"Kamal Nigam, Andrew Kachites Mc-Callum, Sebastian Thrun, Tom Mitchell","year":"2000","venue":"Machine Learning","doi":null,"url":null,"internal_url":null},{"position":22,"raw":"[Pan and Yang, 2010] Sinno Jialin Pan and Qiang Yang. A survey on transfer learning. TKDE, 2010.","title":"A survey on transfer learning","authors":"Jialin Sinno, Qiang Pan, Yang","year":"2010","venue":"TKDE","doi":null,"url":null,"internal_url":null},{"position":23,"raw":"[Rubin, 2004] Donald B Rubin. Multiple imputation for non- response in surveys. John Wiley & Sons, 2004.","title":"Multiple imputation for nonresponse in surveys","authors":"Donald B Rubin","year":"2004","venue":null,"doi":null,"url":null,"internal_url":null},{"position":24,"raw":"[Shimodaira, 2000] Hidetoshi Shimodaira. Improving pre- dictive inference under covariate shift by weighting the log-likelihood function. JSPI, 2000.","title":"Improving predictive inference under covariate shift by weighting the log-likelihood function","authors":"Hidetoshi Shimodaira","year":"2000","venue":"JSPI","doi":null,"url":null,"internal_url":null},{"position":25,"raw":"[Sugiyama et al., 2008] Masashi Sugiyama, Shinichi Naka- jima, Hisashi Kashima, Paul V Buenau, and Motoaki Kawanabe. Direct importance estimation with model se- lection and its application to covariate shift adaptation. In NIPS, 2008.","title":"Direct importance estimation with model selection and its application to covariate shift adaptation","authors":"Sugiyama","year":"2008","venue":"NIPS","doi":null,"url":null,"internal_url":null},{"position":26,"raw":"[Sugiyama et al., 2013] Masashi Sugiyama, Makoto Ya- mada, and Marthinus Christoffel du Plessis. Learning under nonstationarity: covariate shift and class-balance change. Wiley Interdisciplinary Reviews: Computational Statics, 2013.","title":"Learning under nonstationarity: covariate shift and class-balance change","authors":"Masashi Sugiyama, Makoto Yamada, Marthinus Christoffel Du Plessis","year":"2013","venue":"Wiley Interdisciplinary Reviews: Computational Statics","doi":null,"url":null,"internal_url":null},{"position":27,"raw":"[Van der Maaten and Hinton, 2008] Laurens Van der Maaten and Geoffrey Hinton. Visualizing data using t-sne. JMLR, 2008.","title":"Visualizing data using t-sne","authors":"Van Der Maaten, Laurens Hinton, Geoffrey Van Der Maaten, Hinton","year":"2008","venue":null,"doi":null,"url":null,"internal_url":null},{"position":28,"raw":"[Wang et al., 2003] Haixun Wang, Wei Fan, Philip S Yu, and Jiawei Han. Mining concept-drifting data streams using ensemble classifiers. In KDD, 2003.","title":"Mining concept-drifting data streams using ensemble classifiers","authors":"Haixun Wang, Wei Fan, Philip S Yu, Jiawei Han","year":"2003","venue":"KDD","doi":null,"url":null,"internal_url":null},{"position":29,"raw":"[Zhang et al., 2013] Kai Zhang, Vincent Wenchen Zheng, Qiaojun Wang, James Tin-Yau Kwok, Qiang Yang, and Ivan Marsic. Covariate shift in hilbert space: a solution via sorrogate kernels. In ICML, 2013.","title":"Covariate shift in hilbert space: a solution via sorrogate kernels","authors":"Kai Zhang, Vincent Wenchen Zheng, Qiaojun Wang, James Tin-Yau Kwok, Qiang Yang, Ivan Marsic","year":"2013","venue":"ICML","doi":null,"url":null,"internal_url":null}],"license_note":"Copyright International Joint Conferences on Artificial Intelligence."}