Abstract
Multi-output regression aims at learning a mapping from feature variables to multiple output variables. It is significant to utilize variety of inherent relational structure information of observations to conduct multi-output regression task when learning a best mapping from high-dimensional data. In this paper, we propose a new multi-output regression method, which simultaneously takes advantage of the low-rank constraint, sample selection, and feature selection in a unified framework. We first take the effect of low-rank constraint to search the correlation of output variables and impose ℓ 2,p -norm regularization on the coefficient matrix to capture the correlation between features and outputs. And then, the ℓ 2,p -norm on the loss function is designed to discover the correlation between samples, so as to select those informative samples to learn the model for improving predictive capacity. Thirdly, orthogonal subspace learning is exploited to ensure multi-output variables share the same low-rank structure of data by rotating the results of feature selection. In addition, to get the optimal solution of the objective function, we propose an effective iterative optimization algorithm. Finally, we conduct sets of experimental results on real datasets, and show the proposed method outperforms the state-of-the-art methods in terms of aCC and aRMSE.
Similar content being viewed by others
References
Anderson T (1951) Estimating linear restrictions on regression coefficients for multivariate normal distributions. Ann Math Stat 22(3):327–351
Argyriou A, Evgeniou T, Pontil M (2006) Multi-task feature learning. Adv Neural Inf Process Syst 41–48
Bache K, Lichman M UCI machine learning repository. http://archive.ics.uci.edu/ml, University of California, Irvine, School of Information and Computer Sciences
Borchani H, Varando G, Bielza C et al (2015) A survey on multi-output regression. Data Min Knowl 5(5):216–233
Cai X, Ding C, Nie F (2013) On the equivalent of low-rank regressions and linear discriminant analysis based regressions. In: Proceedings of the 19th ACM SIGKDD, pp 1124–1132
Cai X, Nie F, Cai W et al (2013) New graph structured sparsity model for multi-label image annotations. In: ICCV, pp 801–808
Candes EJ, Recht B (2009) Exact matrix completion via convex optimization. Found Comput Math 9(6):717–772
Chang X, Nie F, Yang Y et al (2014) A convex formulation for semi-supervised multi-label feature selection. In: Proceedings of the Twenty-Eighth AAAI Conference on Artificial Intelligence, pp 1171–1177
Chen B, Liu G, Huang Z et al (2011) Multi-task low-rank affinities pursuit for image segmentation. In: Proc. IEEE Int’l Conf. Computer Vision, pp 2439–2446
Dzeroski S, Demsar D, Grbovic J (2000) Predicting chemical parameters of river water quality from bioindicator data. Appl Intell 13(1):7–17
Feng J, Zhou L, Xu H et al (2014) Robust subspace segmentation with block-diagonal prior. In: CVPR, pp 3818–3825
Gao L, Song J et al (2015) Learning in high-dimensional multimedia data: the state of the art. J Multimed Syst 1–11
Gao L, Song J et al (2015) Optimal graph learning with partial tags and multiple features for image and video annotation. In: CVPR, pp 4371–4379
Gao L, Song J et al (2016) Graph-without-cut: an ideal graph learning for image segmentation. In: AAAI Conference on Artificial Intelligence, pp 1188–1194
Gower JC, Dijksterhuis GB (2004) Procrustes problems. Oxford Statistical Science Series, 30. Oxford, UK: Oxford University Press
Izenman AJ (1975) Reduced-rank regression for the multivariate linear model. J Multivar Anal 5(2):248–264
Karalic A, Bratko I (1997) First order regression. Mach Learn 26(2):147–176
Kocev D, Dzeroski S, White MD et al (2009) Using single-target and multi-target regression trees and ensembles to model a compound index of vegetation condition. Ecol Model 220(8):1159–1168
Nie F, Huang H, Cai X et al (2010) Efficient and robust feature selection via joint L21-norms minimization. In: Proc. NIPS, pp 1813–1821
Qin Y, Zhang S, Zhu X et al (2007) Semi-parametric optimization for missing data imputation. Appl Intell 27(1):79–88
Rai P, Kumar A, Daumé H III (2012) Simultaneously leveraging output and task structures for multiple-output regression. Adv Neural Inf Proces Syst 25:1–9
Rothman AJ, Levina E, Zhu J (2010) Sparse multivariate regression with covariance estimation. J Comput Graph Stat 19(4):947–962
Spyromitros-Xioufis E, Tsoumakas G, Groves W et al (2016) Multi-target regression via input space expansion: treating targets as inputs. Mach Learn 1–44
Spyromitros-Xious E, Groves W, Tsoumakas G et al (2012) Multi-label classification methods for multi-target regression. arXiv preprint arXiv:1211.6581, Cornell University Library, pp 1159–1168
Tuia D, Verrelst J, Alonso L et al (2011) Multioutput support vector regression for remote sensing biophysical parameter estimation. IEEE Geosci Remote Sens Lett 8(4):804–808
Wang S, Chang X, Li X, Sheng Q, Chen W et al (2016) Multi-task support vector machines for feature selection with shared knowledge discovery. Signal Process 120:746–753
Wang H, Nie F, Huang H et al (2011) Sparse multi-task regression and feature selection to identify brain imaging predictors for memory performance. In: Proc IEEE Int Conf Comput Vis. pp 557–562. doi:10.1109/ICCV.2011.6126288
Wu F, Yuan Y, Zhuang Y (2010) Heterogeneous feature selection by group lasso with logistic regression. In: ACM MM, pp 983–986
Zhang M, Ding C, Zhang Y et al (2014) Feature selection at the discrete limit. In: Twenty-Eighth AAAI Conference on Artificial Intelligence, pp 1355–1361
Zhang S, Qin Z, Ling CX et al (2005) “Missing is useful”: missing values in cost-sensitive decision trees. IEEE Trans Knowl Data Eng 17(12):1689–1693
Zhao Y, Zhang S (2006) Generalized dimension-reduction framework for recent-biased time series analysis. IEEE Trans Knowl Data Eng 18(2):231–244
Zhu X, Huang Z, Cheng H et al (2013) Sparse hashing for fast multimedia search. ACM Trans Inf Syst 31(2):9.1–9.24
Zhu X, Huang Z, Cheng H et al (2013) Sparse hasing for fast multimedia search. ACM Trans Inf Syst 31(2)
Zhu X, Huang Z, Shen HT et al (2012) Dimensionality reduction by mixed kernel canonical correlation analysis. Pattern Recogn 45(8):3003–3016
Zhu X, Li X, Zhang S et al (2016) Robust joint graph sparse coding for unsupervised spectral feature selection. IEEE Transactions on Neural Networks and Learning Systems, PP(99), pp 1–13
Zhu X, Li X, Zhang S (2015) Block-row sparse multiview multilabel learning for image classification. IEEE Trans Cybern 46(2):450–461
Zhu X, Suk HI, Lee SW et al (2015) Subspace regularized sparse multi-task learning for multi-class neurodegenerative disease identification. IEEE Trans Biomed Eng 63(3):607–618
Zhu X, Zhang S, Jin Z, Zhang Z, Zuoming X (2011) Missing value estimation for mixed-attribute datasets. IEEE Trans Knowl Data Eng 23(1):110–121
Zhu P, Zuo W, Zhang L et al (2015) Unsupervised feature selection by regularized self-representation. Pattern Recogn 48(2):438–446
Acknowledgments
This work was supported in part by the China “1000-Plan” National Distinguished Professorship; the National Natural Science Foundation of China (Grant Nos: 61450001, 61263035, 61573270 and 61672177); the China 973 Program (Grant No: 2013CB329404); the China Key Research Program (Grant No: 2016YFB1000905); the Guangxi Natural Science Foundation (Grant Nos: 2012GXNSFGA060004 and 2015GXNSFCB139011); the Innovation Project of Guangxi Graduate Education (Grant Nos: YCSZ2016046 and YCSZ2016045); the Guangxi Higher Institutions’ Program of Introducing 100 High-Level Overseas Talents; the Guangxi Collaborative Innovation Center of Multi-Source Information Integration and Intelligent Processing; and the Guangxi Bagui Scholar Teams for Innovation and Research Project.
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
About this article
Cite this article
Zhang, S., Yang, L., Deng, Z. et al. Leverage triple relational structures via low-rank feature reduction for multi-output regression. Multimed Tools Appl 76, 17461–17477 (2017). https://doi.org/10.1007/s11042-016-3980-3
Received:
Revised:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s11042-016-3980-3