基于标签特定特征的多目标回归稀疏集成方法

doi:10.3969/j.issn.0372-2112.2020.05.010

电子学报 ›› 2020, Vol. 48 ›› Issue (5): 906-913.DOI: 10.3969/j.issn.0372-2112.2020.05.010

基于标签特定特征的多目标回归稀疏集成方法

刘洪涛^1,2, 李航¹, 王进¹, 李鸽鸽^1,2

1. 重庆邮电大学计算机学院, 重庆 400065;
2. 重庆邮电大学网络智能与网络技术研究中心, 重庆 400065

收稿日期:2019-03-27 修回日期:2019-11-08 出版日期:2020-05-25
- 作者简介:
- 刘洪涛 男,1974年9月生于重庆市北碚,博士,副教授,主要研究方向为社会网络、网络舆情演化、机器学习.E-mail:liuht@cqupt.edu.cn;李航女,1995年生,硕士研究生,主要研究方向为机器学习与数据挖掘.;王进男,1979年生,工学博士,教授,主要研究方向行业大数据分析及系统架构、大规模数据挖掘与机器学习等.;李鸽鸽 女,1995年生,硕士研究生,主要研究方向是机器学习与社会网络.
- 基金资助:
- 国家自然科学基金 (No.41571401）; 国家社会科学基金 (No.18BGL266）

Multi-Target Regression via Sparse Integration and Label-Specific Features

LIU Hong-tao^1,2, LI Hang¹, WANG Jin¹, LI Ge-ge^1,2

1. School of Computer, Chongqing University of Posts and Telecommunications, Chongqing 400065, China;
2. Chongqing University of Posts and Telecommunications Network Intelligence and Network Technology Research Center, Chongqing 400065, China

Received:2019-03-27 Revised:2019-11-08 Online:2020-05-25 Published:2020-05-25

摘要/Abstract

摘要： 多目标回归学习是指同时学习多个相关的回归任务，其主要挑战来自于对输入要素和输出目标变量之间的基础关系进行建模以及对目标间的相关性进行探索.针对这两个挑战，本文提出了一种基于标签特定特征的多目标回归稀疏集成方法，通过探索目标间的相关性，为每个目标构建其独特的标签特定特征，提高算法整体的预测精度；同时设计一种稀疏性聚合函数对不同的回归方法进行集成，从而处理输入与输出间的复杂关系.在18个数据集上与有代表性的多目标回归方法进行对比实验，充分证明了本文方法的有效性与竞争性.

关键词: 多目标回归, 稀疏集成, 标签特定特征, 目标间关联

Abstract: Multi-target regression (MTR) refers to learning multiple relevant regression tasks simultaneously.The main challenges of multi-target regression arise from modeling the underlying relationships between input features and output target variables as well as exploring inter-target correlations.In this paper,we propose a multi-target regression method via sparse integration and label-specific features (SI-LSF) that utilizes inter-target correlations to improve the overall prediction accuracy by constructing label-specific features and deals with the input-output relationships through sparse integration of various regression models.Extensive experimental evaluation on 18 benchmark datasets demonstrates that our proposed method can achieve competitive performance against representative state-of-the-art multi-target regression methods,which shows the great effectiveness in dealing with multivariate prediction.

Key words: multi-target regression, sparse integration, label specific features, inter-target correlations

中图分类号:

TP391
TP399

刘洪涛, 李航, 王进, 李鸽鸽. 基于标签特定特征的多目标回归稀疏集成方法[J]. 电子学报, 2020, 48(5): 906-913.

LIU Hong-tao, LI Hang, WANG Jin, LI Ge-ge. Multi-Target Regression via Sparse Integration and Label-Specific Features[J]. Acta Electronica Sinica, 2020, 48(5): 906-913.

参考文献

[1] Martin B,Dragi K,Saso D.Ensembles for multi-target regression with random output selections[J].Machine Learning,2018,107(11):1673-1709.
[2] Hadavandi E,Shahrabi J,Shamshirband S.A novel boosted-neural network ensemble for modeling multi-target regression problems[J].Engineering Applications of Artificial Intelligence,2015,45(1):204-219.
[3] Yan Y,Ricci E,Subramanian R,et al.Amulti-task learning framework for head pose estimation under target motion[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2016,38(6):1070-1083.
[4] Spyromitros-Xioufis E,Tsoumakas G,Groves W,et al.Multi-target regression via input space expansion:treating targets as inputs[J].Machine Learning,2016,104(1):55-98.
[5] Borchani H,Varando G,Bielza C,et al.A survey on multi-output regression[J].Wiley Interdisciplinary Reviews:Data Mining and Knowledge Discovery,2015,5(5):216-233.
[6] Tsoumakas G,Spyromitros-Xioufis E,Vrekou A,et al.Multi-target regression via random linear target combinations[A].Proceedings of the 2014 Machine Learning and Knowledge Discovery in Databases[C].Nancy:Springer,2014.225-240.
[7] Borchani H,Varando G,Bielza C,et al.A survey on multi-output regression[J].Wiley Interdisciplinary Reviews:Data Mining and Knowledge Discovery,2015,5(5):216-233.
[8] Wang X,Bi J,Yu S,et al.Multiplicativemultitask feature learning[J].Journal of Machine Learning Research,2016,17(1):2820-2852.
[9] Chen J,Liu J,Ye J.Learning incoherent sparse and low-rank patterns from multiple tasks[J].ACM Transactions on Knowledge Discovery from Data,2012,5(4):1-31.
[10] Zhou Q,Zhao Q.Flexibleclustered multi-task learning by learning representative tasks[J].IEEE Transactions on Pattern Analysis & Machine Intelligence,2016,38(2):266-278.
[11] 张晶.基于AdaBoost回归树的多目标预测算法的研究[D].北京:北京交通大学,2017.
[12] Rothman A J,Levina E,Zhu J.Sparsemultivariate regression with covariance estimation[J].Journal of Computational and Graphical Statistics,2010,19(4):947-962.
[13] Zhang Y,Yeung D Y.Aconvex formulation for learning task relationships in multi-task learning[A].Proceedings of the 26th Conference on Uncertainty in Artificial Intelligence[C].Catalina Island:AUAI press,2010.733-742.
[14] Liu H,Wang L,Zhao T.Calibratedmultivariate regression with application to neural semantic basis discovery[J].Journal of Machine Learning Research,2015,16(1):1579-1606.
[15] Alvarez,Mauricio A,Rosasco L,et al.Kernels for vector-valued functions:a review[J].Foundations and Trends in Machine Learning,2011,4(3):195-266.
[16] Jacob L,Bach F,Vert J P.Clustered multi-task learning:A convex formulation[A].Proceedings of the 22nd Annual Conference on Neural Information Processing Systems[C].Vancouver:Curran Associates,2008.745-752.
[17] Yu C,Yao W X.Robust linear regression:a review and comparison[J].Communications in Statistics-Simulation and Computation,2017,46(8):6261-6282.
[18] Chang C C,Lin C J.A library for support vector machines[J].ACM Transactions on Intelligent Systems and Technology,2011,2(3):1-27.
[19] Cutler A,Cutler D R,Stevens J R.Randomforests[J].Machine Learning,2004,45(1):157-176.
[20] Melki G,Cano A,Kecman V,et al.Multi-target support vector regression via correlation regressor chains[J].Information Sciences,2017,415-416(1):53-69.
[21] Demšar J.Statistical comparisons of classifiers over multiple data sets[J].Journal of Machine Learning Research,2006,7(1):1-30.

基于标签特定特征的多目标回归稀疏集成方法

Multi-Target Regression via Sparse Integration and Label-Specific Features

PDF

可视化

摘要/Abstract

引用本文

使用本文

参考文献

相关文章 1

编辑推荐

Metrics