{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2025,2,21]],"date-time":"2025-02-21T14:27:30Z","timestamp":1740148050817,"version":"3.37.3"},"reference-count":52,"publisher":"Association for Computing Machinery (ACM)","issue":"4","funder":[{"name":"Chongqing Overseas Chinese Entrepreneurship and Innovation Support Program and the National Science Foundation of China","award":["62176055"]}],"content-domain":{"domain":["dl.acm.org"],"crossmark-restriction":true},"short-container-title":["ACM Trans. Knowl. Discov. Data"],"published-print":{"date-parts":[[2024,5,31]]},"abstract":"\n Multiple-instance learning<\/jats:italic>\n (MIL) solves the problem where training instances are grouped in bags, and a binary (positive or negative) label is provided for each bag. Most of the existing MIL studies need fully labeled bags for training an effective classifier, while it could be quite hard to collect such data in many real-world scenarios, due to the high cost of data labeling process. Fortunately, unlike fully labeled data,\n triplet comparison data<\/jats:italic>\n can be collected in a more accurate and human-friendly way. Therefore, in this article, we for the first time investigate MIL from\n only triplet comparison bags<\/jats:italic>\n , where a triplet\n \n (X\n a<\/jats:sup>\n , X\n b<\/jats:sup>\n , X\n c<\/jats:sup>\n )\n <\/jats:italic>\n contains the weak supervision information that bag\n \n X\n a<\/jats:sup>\n <\/jats:italic>\n is more similar to\n \n X\n b<\/jats:sup>\n <\/jats:italic>\n than to\n \n X\n c<\/jats:sup>\n <\/jats:italic>\n . To solve this problem, we propose to train a bag-level classifier by the empirical risk minimization framework and theoretically provide a generalization error bound. We also show that a convex formulation can be obtained only when specific convex binary losses such as the square loss and the double hinge loss are used. Extensive experiments validate that our proposed method significantly outperforms other baselines.\n <\/jats:p>","DOI":"10.1145\/3638776","type":"journal-article","created":{"date-parts":[[2024,1,2]],"date-time":"2024-01-02T21:58:31Z","timestamp":1704232711000},"page":"1-18","update-policy":"https:\/\/doi.org\/10.1145\/crossmark-policy","source":"Crossref","is-referenced-by-count":1,"title":["Multiple-instance Learning from Triplet Comparison Bags"],"prefix":"10.1145","volume":"18","author":[{"ORCID":"https:\/\/orcid.org\/0009-0008-7861-4834","authenticated-orcid":false,"given":"Senlin","family":"Shu","sequence":"first","affiliation":[{"name":"Chongqing University, China"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-6130-7220","authenticated-orcid":false,"given":"Deng-Bao","family":"Wang","sequence":"additional","affiliation":[{"name":"Southeast University, China"}]},{"ORCID":"https:\/\/orcid.org\/0009-0008-3747-0635","authenticated-orcid":false,"given":"Suqin","family":"Yuan","sequence":"additional","affiliation":[{"name":"The University of Sydney, Australia"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-8973-2843","authenticated-orcid":false,"given":"Hongxin","family":"Wei","sequence":"additional","affiliation":[{"name":"Southern University of Science and Technology, China"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-8249-1725","authenticated-orcid":false,"given":"Jiuchuan","family":"Jiang","sequence":"additional","affiliation":[{"name":"Nanjing University of Finance and Economics, China"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-2839-5799","authenticated-orcid":false,"given":"Lei","family":"Feng","sequence":"additional","affiliation":[{"name":"Nanyang Technological University, Singapore"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-1880-5918","authenticated-orcid":false,"given":"Min-Ling","family":"Zhang","sequence":"additional","affiliation":[{"name":"Southeast University, China"}]}],"member":"320","published-online":{"date-parts":[[2024,2,12]]},"reference":[{"key":"e_1_3_5_2_2","doi-asserted-by":"publisher","DOI":"10.1016\/j.artint.2013.06.003"},{"key":"e_1_3_5_3_2","article-title":"CVXOPT: Python software for convex optimization","author":"Andersen Martin S.","year":"2013","unstructured":"Martin S. Andersen, Joachim Dahl, and Lieven Vandenberghe. 2013. CVXOPT: Python software for convex optimization. Retrieved from https:\/\/cvxopt.org","journal-title":"Retrieved from https:\/\/cvxopt.org"},{"key":"e_1_3_5_4_2","first-page":"577","volume-title":"Proceedings of the NeurIPS","author":"Andrews Stuart","year":"2002","unstructured":"Stuart Andrews, Ioannis Tsochantaridis, and Thomas Hofmann. 2002. Support vector machines for multiple-instance learning. In Proceedings of the NeurIPS. 577\u2013584."},{"key":"e_1_3_5_5_2","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2009.5206737"},{"key":"e_1_3_5_6_2","first-page":"452","volume-title":"Proceedings of the ICML","author":"Bao Han","year":"2018","unstructured":"Han Bao, Gang Niu, and Masashi Sugiyama. 2018. Classification from pairwise similarity and unlabeled data. In Proceedings of the ICML. 452\u2013461."},{"key":"e_1_3_5_7_2","doi-asserted-by":"publisher","DOI":"10.1016\/j.neunet.2018.05.001"},{"issue":"11","key":"e_1_3_5_8_2","first-page":"463","article-title":"Rademacher and Gaussian complexities: Risk bounds and structural results","volume":"3","author":"Bartlett Peter L.","year":"2002","unstructured":"Peter L. Bartlett and Shahar Mendelson. 2002. Rademacher and Gaussian complexities: Risk bounds and structural results. J. Mach. Learn. Res. 3, 11 (2002), 463\u2013482.","journal-title":"J. Mach. Learn. Res."},{"key":"e_1_3_5_9_2","first-page":"1272","volume-title":"Proceedings of the ICML","author":"Cao Yuzhou","year":"2021","unstructured":"Yuzhou Cao, Lei Feng, Yitian Xu, Bo An, Gang Niu, and Masashi Sugiyama. 2021. Learning from similarity-confidence data. In Proceedings of the ICML. 1272\u20131282."},{"key":"e_1_3_5_10_2","doi-asserted-by":"publisher","DOI":"10.1016\/j.patcog.2017.10.009"},{"key":"e_1_3_5_11_2","doi-asserted-by":"publisher","DOI":"10.1109\/TPAMI.2006.248"},{"key":"e_1_3_5_12_2","article-title":"Deep learning for classical japanese literature","author":"Clanuwat Tarin","year":"2018","unstructured":"Tarin Clanuwat, Mikel Bober-Irizar, Asanobu Kitamoto, Alex Lamb, Kazuaki Yamamoto, and David Ha. 2018. Deep learning for classical japanese literature. Retrieved from https:\/\/arXiv:1812.01718","journal-title":"Retrieved from https:\/\/arXiv:1812.01718"},{"key":"e_1_3_5_13_2","doi-asserted-by":"publisher","DOI":"10.1162\/neco_a_01262"},{"key":"e_1_3_5_14_2","doi-asserted-by":"publisher","DOI":"10.1016\/S0004-3702(96)00034-3"},{"key":"e_1_3_5_15_2","first-page":"1386","volume-title":"Proceedings of the ICML","author":"Plessis M. C. du","year":"2015","unstructured":"M. C. du Plessis, G. Niu, and M. Sugiyama. 2015. Convex formulation for learning from positive and unlabeled data. In Proceedings of the ICML. 1386\u20131394."},{"key":"e_1_3_5_16_2","doi-asserted-by":"publisher","DOI":"10.1145\/3447548.3467318"},{"key":"e_1_3_5_17_2","doi-asserted-by":"publisher","DOI":"10.1017\/S026988890999035X"},{"key":"e_1_3_5_18_2","first-page":"179","volume-title":"Proceedings of the ICML","author":"G\u00e4rtner Thomas","year":"2002","unstructured":"Thomas G\u00e4rtner, Peter A. Flach, Adam Kowalczyk, and Alexander J. Smola. 2002. Multi-instance kernels. In Proceedings of the ICML. 179\u2013186."},{"key":"e_1_3_5_19_2","volume-title":"Proceedings of the NeurIPS","author":"Han Bo","year":"2018","unstructured":"Bo Han, Quanming Yao, Xingrui Yu, Gang Niu, Miao Xu, Weihua Hu, Ivor Tsang, and Masashi Sugiyama. 2018. Co-teaching: Robust training of deep neural networks with extremely noisy labels. In Proceedings of the NeurIPS."},{"key":"e_1_3_5_20_2","doi-asserted-by":"publisher","DOI":"10.1109\/TPAMI.2018.2861732"},{"key":"e_1_3_5_21_2","first-page":"2127","volume-title":"Proceedings of the ICML","author":"Ilse Maximilian","year":"2018","unstructured":"Maximilian Ilse, Jakub Tomczak, and Max Welling. 2018. Attention-based deep multiple instance learning. In Proceedings of the ICML. PMLR, 2127\u20132136."},{"key":"e_1_3_5_22_2","first-page":"5917","volume-title":"Proceedings of the NeurIPS","author":"Ishida T.","year":"2018","unstructured":"T. Ishida, G. Niu, and M. Sugiyama. 2018. Binary classification for positive-confidence data. In Proceedings of the NeurIPS. 5917\u20135928."},{"key":"e_1_3_5_23_2","doi-asserted-by":"publisher","DOI":"10.1109\/5.726791"},{"key":"e_1_3_5_24_2","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-642-15567-3_3"},{"key":"e_1_3_5_25_2","first-page":"1","article-title":"Deep multiple instance selection","volume":"64","author":"Li Xin-Chun","year":"2021","unstructured":"Xin-Chun Li, De-Chuan Zhan, Jia-Qi Yang, and Yi Shi. 2021. Deep multiple instance selection. Sci. China Info. Sci. 64 (2021), 1\u201315.","journal-title":"Sci. China Info. Sci."},{"key":"e_1_3_5_26_2","doi-asserted-by":"publisher","DOI":"10.1016\/j.neucom.2020.12.009"},{"key":"e_1_3_5_27_2","volume-title":"Proceedings of the ICLR","author":"Lu Nan","year":"2019","unstructured":"Nan Lu, Gang Niu, Aditya K. Menon, and Masashi Sugiyama. 2019. On the minimal supervision for training any binary classifier from only unlabeled data. In Proceedings of the ICLR."},{"key":"e_1_3_5_28_2","first-page":"281","volume-title":"Proceedings of the BSMSP","author":"MacQueen James","year":"1967","unstructured":"James MacQueen et\u00a0al. 1967. Some methods for classification and analysis of multivariate observations. In Proceedings of the BSMSP. 281\u2013297."},{"issue":"1","key":"e_1_3_5_29_2","first-page":"148","article-title":"On the method of bounded differences","volume":"141","author":"McDiarmid Colin","year":"1989","unstructured":"Colin McDiarmid. 1989. On the method of bounded differences. Surveys Combinator. 141, 1 (1989), 148\u2013188.","journal-title":"Surveys Combinator."},{"key":"e_1_3_5_30_2","doi-asserted-by":"publisher","DOI":"10.1109\/TIT.2008.926323"},{"key":"e_1_3_5_31_2","doi-asserted-by":"publisher","DOI":"10.5555\/2371238"},{"key":"e_1_3_5_32_2","first-page":"1199","volume-title":"Proceedings of the NeurIPS","author":"Niu Gang","year":"2016","unstructured":"Gang Niu, Marthinus Christoffel du Plessis, Tomoya Sakai, Yao Ma, and Masashi Sugiyama. 2016. Theoretical comparisons of positive-unlabeled learning against positive-negative learning. In Proceedings of the NeurIPS. 1199\u20131207."},{"key":"e_1_3_5_33_2","doi-asserted-by":"publisher","DOI":"10.1186\/1471-2105-6-S1-S18"},{"key":"e_1_3_5_34_2","doi-asserted-by":"publisher","DOI":"10.1145\/1102351.1102439"},{"key":"e_1_3_5_35_2","first-page":"2998","volume-title":"Proceedings of the ICML","author":"Sakai Tomoya","year":"2017","unstructured":"Tomoya Sakai, Marthinus Christoffel du Plessis, Gang Niu, and Masashi Sugiyama. 2017. Semi-supervised classification based on classification from positive and unlabeled data. In Proceedings of the ICML. 2998\u20133006."},{"key":"e_1_3_5_36_2","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2015.7298682"},{"key":"e_1_3_5_37_2","article-title":"Classification from pairwise similarities\/dissimilarities and unlabeled data via empirical risk minimization","author":"Shimada Takuya","year":"2020","unstructured":"Takuya Shimada, Han Bao, Issei Sato, and Masashi Sugiyama. 2020. Classification from pairwise similarities\/dissimilarities and unlabeled data via empirical risk minimization. Neural Comput. 33, 5 (2020), 1234\u20131268.","journal-title":"Neural Comput."},{"key":"e_1_3_5_38_2","doi-asserted-by":"publisher","DOI":"10.1145\/1015330.1015405"},{"key":"e_1_3_5_39_2","first-page":"577","volume-title":"Proceedings of the ICML","author":"Wagstaff Kiri","year":"2001","unstructured":"Kiri Wagstaff, Claire Cardie, Seth Rogers, Stefan Schr\u00f6dl et\u00a0al. 2001. Constrained k-means clustering with background knowledge. In Proceedings of the ICML. 577\u2013584."},{"key":"e_1_3_5_40_2","doi-asserted-by":"publisher","DOI":"10.1137\/1.9781611972788.15"},{"key":"e_1_3_5_41_2","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR42600.2020.01374"},{"key":"e_1_3_5_42_2","article-title":"Fashion-MNIST: A novel image dataset for benchmarking machine learning algorithms","author":"Xiao Han","year":"2017","unstructured":"Han Xiao, Kashif Rasul, and Roland Vollgraf. 2017. Fashion-MNIST: A novel image dataset for benchmarking machine learning algorithms. Retrieved from https:\/\/arXiv:1708.07747","journal-title":"Retrieved from https:\/\/arXiv:1708.07747"},{"issue":"1","key":"e_1_3_5_43_2","article-title":"Sensor localization based on nonmetric multidimensional scaling","volume":"2","author":"Xiao Ling","year":"2006","unstructured":"Ling Xiao, Renfa Li, Juan Luo, et\u00a0al. 2006. Sensor localization based on nonmetric multidimensional scaling. STRESS 2, 1 (2006).","journal-title":"STRESS"},{"key":"e_1_3_5_44_2","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-540-24775-3_35"},{"key":"e_1_3_5_45_2","first-page":"1681","article-title":"Multiple-instance pruning for learning efficient cascade detectors","author":"Zhang Cha","year":"2007","unstructured":"Cha Zhang and Paul Viola. 2007. Multiple-instance pruning for learning efficient cascade detectors. In Proceedings of the NeurIPS. 1681\u20131688.","journal-title":"Proceedings of the NeurIPS"},{"key":"e_1_3_5_46_2","doi-asserted-by":"publisher","DOI":"10.1109\/TKDE.2017.2721942"},{"key":"e_1_3_5_47_2","first-page":"1073","volume-title":"Proceedings of the NeurIPS","author":"Zhang Qi","year":"2001","unstructured":"Qi Zhang and Sally A. Goldman. 2001. EM-DD: An improved multiple-instance learning technique. In Proceedings of the NeurIPS. 1073\u20131080."},{"key":"e_1_3_5_48_2","doi-asserted-by":"publisher","DOI":"10.24963\/ijcai.2020\/330"},{"key":"e_1_3_5_49_2","first-page":"34940","article-title":"Multi-instance causal representation learning for instance label prediction and out-of-distribution generalization","author":"Zhang Weijia","year":"2022","unstructured":"Weijia Zhang, Xuanhui Zhang, Min-Ling Zhang et\u00a0al. 2022. Multi-instance causal representation learning for instance label prediction and out-of-distribution generalization. In Proceedings of the NeurIPS. 34940\u201334953.","journal-title":"Proceedings of the NeurIPS"},{"key":"e_1_3_5_50_2","article-title":"Multi-instance multi-label learning with application to scene classification","author":"Zhang Zhi-Li","year":"2006","unstructured":"Zhi-Li Zhang and Min-Ling Zhang. 2006. Multi-instance multi-label learning with application to scene classification. In Proceedings of the NeurIPS.","journal-title":"Proceedings of the NeurIPS"},{"key":"e_1_3_5_51_2","doi-asserted-by":"publisher","DOI":"10.1093\/nsr\/nwx106"},{"key":"e_1_3_5_52_2","first-page":"1249","volume-title":"Proceedings of the ICML","author":"Zhou Zhi-Hua","year":"2009","unstructured":"Zhi-Hua Zhou, Yu-Yin Sun, and Yu-Feng Li. 2009. Multi-instance learning by treating instances as non-iid samples. In Proceedings of the ICML. 1249\u20131256."},{"issue":"1","key":"e_1_3_5_53_2","first-page":"1","article-title":"Introduction to semi-supervised learning","volume":"3","author":"Zhu Xiaojin","year":"2009","unstructured":"Xiaojin Zhu and Andrew B. Goldberg. 2009. Introduction to semi-supervised learning. Synth. Lect. Artific. Intell. Mach. Learn. 3, 1 (2009), 1\u2013130.","journal-title":"Synth. Lect. Artific. Intell. Mach. Learn."}],"container-title":["ACM Transactions on Knowledge Discovery from Data"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/dl.acm.org\/doi\/pdf\/10.1145\/3638776","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2024,2,12]],"date-time":"2024-02-12T12:09:26Z","timestamp":1707739766000},"score":1,"resource":{"primary":{"URL":"https:\/\/dl.acm.org\/doi\/10.1145\/3638776"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2024,2,12]]},"references-count":52,"journal-issue":{"issue":"4","published-print":{"date-parts":[[2024,5,31]]}},"alternative-id":["10.1145\/3638776"],"URL":"https:\/\/doi.org\/10.1145\/3638776","relation":{},"ISSN":["1556-4681","1556-472X"],"issn-type":[{"type":"print","value":"1556-4681"},{"type":"electronic","value":"1556-472X"}],"subject":[],"published":{"date-parts":[[2024,2,12]]},"assertion":[{"value":"2022-06-08","order":0,"name":"received","label":"Received","group":{"name":"publication_history","label":"Publication History"}},{"value":"2023-12-15","order":1,"name":"accepted","label":"Accepted","group":{"name":"publication_history","label":"Publication History"}},{"value":"2024-02-12","order":2,"name":"published","label":"Published","group":{"name":"publication_history","label":"Publication History"}}]}}