Report title: Conditional probability estimation based classification with class label missing at random
Reporter: Professor Wang Qihua, Institute of Mathematics and Systems Science, Chinese Academy of Sciences
Reporting time: January 11, 2021 10:00-10:50
Report location: Zoom meeting (Zoom meeting id: 770 311 8512, password: 378548)
School contact: Wang Peijie wangpeijie@jlu.edu.cn
Report summary:
For binary classification, it is common that class labels of some subjects are missing. Generally, the complete case analysis and the two stage procedure can be used to extend existing full data classification methods to deal with classification with missing class labels. Nevertheless , these two approaches can not take full advantage of unlabeled subjects. In this paper, binary classification with the class label missing at random (MAR) is considered. Based on the inverse probability weighting (IPW) method and the augmented inverse probability weighting (AIPW ) method, two new methods called IPW-CPC and AIPW-CPC are proposed to construct powerful classifiers by estimating the conditional probability in a reproducing kernel Hilbert space (RKHS). Compared with the complete case analysis and the two stage procedure, the proposed IPW -CPC and AIPW-CPC methods can make the best use of unlabeled subjects, which contributes a lot to improving classification accuracy. Theoretically, we sho w that conditional misclassification rates of the proposed classifiers converge to the Bayes misclassification rate in probability and rates of convergence are also obtained. Finally, simulations and the real data analysis well demonstrate good performances of the proposed IPW-CPC and AIPW-CPC methods in comparison with existing methods.
Speaker's profile:
Wang Qihua, researcher at the Chinese Academy of Mathematics and Systems Science, doctoral supervisor, winner of the National Outstanding Youth Fund, distinguished professor of the Yangtze River Scholars Award Program of the Ministry of Education, candidate of the "Hundred Talents Program" of the Chinese Academy of Sciences, the first national hundred outstanding doctors The winner of the paper award. He taught at Peking University and the University of Hong Kong, and visited Carleton University in Canada, University of California at Davis, University of California at Los Angeles, Yale University, University of Washington, Northwestern University, Humboldt University in Germany, Australian National University and Australia University of Sydney etc. Mainly engaged in research on survival analysis, missing data analysis, high-dimensional data statistical analysis and non-semiparametric statistical inference. Published three monographs and published more than 100 papers in important international journals such as The Annals of Statistics, JASA and Biometrika. He is the chairman of the High-Dimensional Statistics Branch, the vice chairman of the Survival Analysis and Biostatistics Branch, a member of the IMS-China and IBS-China committees, and the editorial board of some international and domestic academic journals.
