Moment-Guided Discriminative Manifold Correlation Learning on Ordinal Data

Qing Tian, Wenqiang Zhang, Meng Cao, Liping Wang, Songcan Chen, Hujun Yin

Research output: Contribution to journalArticlepeer-review

115 Downloads (Pure)

Abstract

Canonical correlation analysis (CCA) is a typical and useful learning paradigm in big data analysis for capturing correlation across multiple views of the same objects. When dealing with such data with additional ordinal information, traditional CCA suffers from poor performance due to ignoring the ordinal relationships within the data. Such data is becoming increasingly common as either temporal or sequential information is often associated with the data collection process. In order to incorporate the ordinal information into the objective function of CCA, the so-called ordinal discriminative CCA (OR-DisCCA) has been presented in the literature. Although OR-DisCCA can yield better ordinal regression results, its performance deteriorates when the data is corrupted with noise and outliers, as they tend to smear the order information contained in class centers. To address this issue, in this paper we construct a robust, manifold-preserved ordinal discriminative correlation regression (rmODCR). The robustness is achieved by replacing the traditional (l2-norm) class centers with lp-norm centers where p is efficiently estimated according to the moments of the data distributions, as well as by incorporating the manifold distribution information of the data in the objective optimization. In addition, we further extend rmODCR with deep convolutional architectures. Extensive experimental evaluations have demonstrated the superiority of the proposed method.
Original languageEnglish
Pages (from-to)1-18
Number of pages18
JournalACM Transactions on Intelligent Systems and Technology
Volume11
Issue number5
DOIs
Publication statusPublished - 25 Jul 2020

Fingerprint

Dive into the research topics of 'Moment-Guided Discriminative Manifold Correlation Learning on Ordinal Data'. Together they form a unique fingerprint.

Cite this