·958· 智能系统学报 第14卷 [7]JAPKOWICZ N,STEPHEN S.The class imbalance prob- [17]LI Feifei,FERGUS R,PERONA P.Learning generative lem:A systematic study[J].Intelligent data analysis,2002, visual models from few training examples:an increment- 6(5):429-449. al Bayesian approach tested on 101 object categories[J]. [8]MORENO-TORRES J G,HERRERA F.A preliminary Computer vision and image understanding,2007,106(1): study on overlapping and data fracture in imbalanced do- 59-70. mains by means of Genetic Programming-based feature ex- [18]RUSSAKOVSKY O,DENG Jia,SU Hao,et al.ImageN- traction[C]//Proceedings of the 201010th International et large scale visual recognition challenge[J].Internation- Conference on Intelligent Systems Design and Applica- al journal of computer vision,2015,115(3):211-252. tions.Cairo,Egypt,2014:501-506. [19]HE Kaiming,ZHANG Xiangyu,REN Shaoqing,et al. [9]WANG K J,MAKOND B,CHEN Kunhuang,et al.A hy- Deep residual learning for image recognition[Cl/Proceed- brid classifier combining SMOTE with PSO to estimate 5- ings of the IEEE conference on computer vision and pat- year survivability of breast cancer patients[J].Applied soft tern recognition.Las Vegas,NV,USA,2016:770-778. computing,2014.20:15-24. [20]ESPINDOLA R P,EBECKEN N FF.On extending F. [10]CHAWLA N V.BOWYER K W,HALL L O,et al. measure and G-mean metrics to multi-class problems[Ml SMOTE:synthetic minority over-sampling technique[J]. ZANASI A.BREBBIA C A.EBECKEN N FF.Data Journal of artificial intelligence research,2002,16(1): Mining VI Data Mining,Text Mining and Their Business 321-357 Applications.Southampton:WIT Press,2005,25-34. [11]KOPLOWITZ J,BROWN T A.On the relation of per- 作者简介： formance to editing in nearest neighbor rules[J].Pattern 黄庆康，男，1994年生，硕士研究 recognition,.1981,133):251-255. 生，主要研究方向为图像分类、广告 [12]CATENI S,COLLA V.VANNUCCI M.A method for 推荐。 resampling imbalanced datasets in binary classification tasks for real-world problems[J].Neurocomputing,2014, 135:32-41. [13]ELKAN C.The foundations of cost-sensitive learning[Cl// Proceedings of the Seventeenth International Joint Con- 宋恺涛，男，1993年生，博士，主 ference on Artificial Intelligence.San Francisco,USA, 要研究方向为数据挖掘、推荐系统。 2001:973-978 [14]ZHOU Zhihua,LIU Xuying.Training cost-sensitive neur- al networks with methods addressing the class imbalance problem[J].IEEE transactions on knowledge and data en- gineering,.2006,18(1):63-77. [15]LIN T Y,GOYAL P,GIRSHICK R,et al.Focal loss for 陆建峰，男，1969年生，教授，主 要研究方向为模式识别。参与过近 dense object detection[J].IEEE transactions on pattern 20项省部级课题，获各类省部级科技 analysis and machine intelligence,2017,1:2999-3007. 进步奖9项。发表学术论文80余篇。 [16]GOODFELLOW I,BENGIO Y,COURVILLE A.Deep learning[M].Cambridge,massachusetts:MIT press,2016: 218-227.JAPKOWICZ N, STEPHEN S. The class imbalance prob￾lem: A systematic study[J]. Intelligent data analysis, 2002, 6(5): 429–449. [7] MORENO-TORRES J G, HERRERA F. A preliminary study on overlapping and data fracture in imbalanced do￾mains by means of Genetic Programming-based feature ex￾traction[C]//Proceedings of the 201010th International Conference on Intelligent Systems Design and Applica￾tions. Cairo, Egypt, 2014: 501−506. [8] WANG K J, MAKOND B, CHEN Kunhuang, et al. A hy￾brid classifier combining SMOTE with PSO to estimate 5- year survivability of breast cancer patients[J]. Applied soft computing, 2014, 20: 15–24. [9] CHAWLA N V, BOWYER K W, HALL L O, et al. SMOTE: synthetic minority over-sampling technique[J]. Journal of artificial intelligence research, 2002, 16(1): 321–357. [10] KOPLOWITZ J, BROWN T A. On the relation of per￾formance to editing in nearest neighbor rules[J]. Pattern recognition, 1981, 13(3): 251–255. [11] CATENI S, COLLA V, VANNUCCI M. A method for resampling imbalanced datasets in binary classification tasks for real-world problems[J]. Neurocomputing, 2014, 135: 32–41. [12] ELKAN C. The foundations of cost-sensitive learning[C]// Proceedings of the Seventeenth International Joint Con￾ference on Artificial Intelligence. San Francisco, USA, 2001: 973−978. [13] ZHOU Zhihua, LIU Xuying. Training cost-sensitive neur￾al networks with methods addressing the class imbalance problem[J]. IEEE transactions on knowledge and data en￾gineering, 2006, 18(1): 63–77. [14] LIN T Y, GOYAL P, GIRSHICK R, et al. Focal loss for dense object detection[J]. IEEE transactions on pattern analysis and machine intelligence, 2017, 1: 2999–3007. [15] GOODFELLOW I, BENGIO Y, COURVILLE A. Deep learning[M]. Cambridge, massachusetts: MIT press, 2016: 218−227. [16] LI Feifei, FERGUS R, PERONA P. Learning generative visual models from few training examples: an increment￾al Bayesian approach tested on 101 object categories[J]. Computer vision and image understanding, 2007, 106(1): 59–70. [17] RUSSAKOVSKY O, DENG Jia, SU Hao, et al. ImageN￾et large scale visual recognition challenge[J]. Internation￾al journal of computer vision, 2015, 115(3): 211–252. [18] HE Kaiming, ZHANG Xiangyu, REN Shaoqing, et al. Deep residual learning for image recognition[C]//Proceed￾ings of the IEEE conference on computer vision and pat￾tern recognition. Las Vegas, NV, USA, 2016: 770−778. [19] ESPÍNDOLA R P, EBECKEN N F F. On extending F￾measure and G-mean metrics to multi-class problems[M]// ZANASI A, BREBBIA C A, EBECKEN N F F. Data Mining VI Data Mining, Text Mining and Their Business Applications. Southampton: WIT Press, 2005, 25−34. [20] 作者简介： 黄庆康，男，1994 年生，硕士研究 生，主要研究方向为图像分类、广告 推荐。 宋恺涛，男，1993 年生，博士，主 要研究方向为数据挖掘、推荐系统。 陆建峰，男，1969 年生，教授，主 要研究方向为模式识别。参与过近 20 项省部级课题，获各类省部级科技 进步奖 9 项。发表学术论文 80 余篇。 ·958· 智 能 系 统 学 报 第 14 卷