Chenping Hou(侯臣平)

Associate Professor, National University of Defense Technology
Research Interests:Machine LearningStatistical data analysisPattern RecognitionComputer Version

Welcome to visit Chenping Hou's HomePage


See the full text in the ResearchGate at  


Visiting Student in Bigeye Lab, Tsinghua University, Supervised by Professor Changshui Zhang, 2008.

Visiting Scholar in QCIS, UTS, Supervised by Professor Dacheng Tao, 2013.

Visiting Scholar in Lamda, Nanjing Univesity, Supervised by Professor Zhi-Hua Zhou, 2016.

Research Interests

Statistical data analysis: high dimensional data analysis, statistical machine learning, pattern recognition, etc.

Applications: Image processing, multiple view data analysis, etc.

Representative Publications

(One can see Publications for papers and codes)

Robust Learning

1.   C. Hou, and Z.-H. Zhou. One-Pass Learning with Incremental and Decremental Features. Accepted by IEEE TPAMI.

We focus on the problem with both feature and instance evolving and propose an one-pass learning method to solve this problem.

2.   C. Hou, L.-L. Zeng, D. Hu, Secure Classification with augmented features. arXiv,

     We propose a secure classification approach, whose accuracy is never degenerated when exploiting augmented features.

Dimensionality Rediction

1.   C. Hou, F.Nie, D. Yi and D. Tao. Discriminative embeded clustering: A general framework for grouping high dimensional data. Accepted by IEEE TNNLS, 2014.

We focus on the unsupervised problem of learning subspace for clustering, which have been employed sequencely in traditional applications. It is a genaral framework that can analyze many methods in a unified view.

2.   C. Hou, C.Zhang, Y. Wu, Y. Jiao. Stable local dimensionality reduction approaches, Pattern Recognition, Volume 42, Issue 9, September 2009, Pages 2054-2066.

We focus on the problem of enhance the stability of local dimensionality reduction approaches. We have proposed a novel framework which can add global information for local methods.

3.   C. Hou, J.Wang, Y. Wu, D. Yi. Local linear transformation embedding, Neurocomputing, Volume 72, Issues 10–12, June 2009, Pages 2368-2378.

We focus on the problem of  alleviate the local sensitivity of LLE. We compute the local approximation weight in a targent space.

4.   C. Hou, F.Nie, C. Zhang, Y. Wu. Learning an orthogonal and smooth subspace for image classification. Signal Processing Letters, IEEE, 16 (4): 303-306, 2009.

We focus on the problem of subspace learning for image classification. We add spatial smooth regularizer and use orthogonal constraint in learning subspace.

5.   C. Hou, F. Nie,C. Zhang, Y. Wu. Learning a Subspace for Clustering via Pattern Shrinking. Information Processing & Management (IPM), 49(4):871-883, 2013.

We focus on the problem of learning subspace for clustering. We propose a nove strategy, named as pattern shinking to maintain the nonlinear structure in linear subspace learning.

Feature Selection

1.   C. Hou, F.Nie, X. Li, D. Yi, Y. Wu. Joint Embedding Learning and Sparse Regression: A   Framework for Unsupervised Feature Selection. IEEE Transactions on Cybernetics,44(6): 793 – 804, 2014.

We focus on the problem of unsupervised feature selection. A general framework which can joint embedding learning and sparse regression has been proposed.  It enables traditional dimensionality reduction method, i.e., spectral embedding, for feature selection by adding the sparse constaints.

2.   C. Hou, F.Nie, D. Yi, Y Wu. Feature selection via joint embedding learning and sparseregression. Proceedings of the Twenty-Second international joint conference on Artificial Intelligence, vol. 2, pp. 1324-1329, 2011.

We focus on the problem of unsupervised feature selection.  It is a short version of the paper "Joint embedding learning and sparse regression: a general framework for unsupervised feature selection".

Multiple Rank Regression

1.   C. Hou, F.Nie, D. Yi, Y. Wu. Efficient image classification via multiple rank regression.IEEE Transactions on Image Processing, 22 (1): 340-352, 2013.

We focus on the problem of image data classification. We propose a novel multiple rank regresion model to classify image data directly. We also reveal the essence of the rank of regresion. It can be regarded as a general extension of traditional linear regression method.

2.   C. Hou, F.Nie, C. Zhang, D. Yi, Y. Wu. Multiple rank multi linear SVM for matrix data classification. Pattern Recognition 47 (1): 454-469, 2014.

We focus on the problem of enabling linear SVM manipulating matrix data directly. A general method which uses multiple rank multiple linear regression as the constraints of linear SVM has been presented.

Weak Supervised Learning

1.   C. Hou, F.Nie, F. Wang, C. Zhang, Y. Wu. Semisupervised Learning Using Negative Labels. IEEETransactions on Neural Networks, 22 (3), Mar. 2011, pp. 420-432.
We focus on the problem of semi supervised learning.  We proposed a new kind of label, i.e., negative label, and also demonstrate its effectiveness in supervised learning.

2.   C. Hou,C.  Zhang, Y. Wu, F. Nie. Multiple view semi-supervised dimensionality reduction, Pattern Recognition, Volume 43, Issue3, March 2010, Pages 720-730.

We focus on the problem of semi supervised dimensionalty reduction on multiple view data.  It is proposed to use link information for multiple view dimensionality reduction.


2017.11.9 Congratulations. Our paper named as Reliable multi-view learning has been accepted by AAAI 2018.

2017.10.22 Congratulations. Our paper named as One-pass learning with incremental and decremental features has been accepted by TPAMI.

2014.7.16 Our new paper for learning subspace for clustering has been accepted by TNNLS.

2014.7.20 We have released the codes for most of our papers.

2014.7.23 We have updated the code for MRR.

2014.10.16 I will give two posters at MLA 14 in Xi'an.

2014.10.16 I will host Fei Wang at IBM to give a talk in 5th, Nov, Wellcome!

2014.12.23 I will give a talk 'Learning low dimensional Structures from high dimensional data' on 'Workshop on Frontiers Mathematical Problems in Network Science and Data Geometry', 8th, Jan, 2015. Wellcome!

2015.04.17 Congratulations, Tao's paper, Effective Discriminative Feature Selection with Non-trival Solutions has been accepted by IEEE Transaction on Neural Networks and Learning System.

2015.04.23 I will give a poster named as multiple rank linear SVM in VALSE 2015@Chengdu. Wellcome!

2015.11.13 Our paper 'Discriminative vanish component analysis' has been accepted by AAAI16.

2016.07.02, Call for papers: special issue on Optimization in Machine Learning and Data Mining of the journal Journal of Optimization, co-organized with Dr. Tongliang Liu and Dr. Feiping Nie.

2016.9.12 Congratulations. Our two papers named 'semi-supvervised multi-label dimensionality reduction' and 'unsupervised feature extraction using a learned graph with clustering structure' have been accepted by ICDM 2016 and ICPR 2016 respectively. Thanks to Baolin Guo and Wenzhang Zhuge.

2017.6.15 Congratulations. Four papers, named as 'mult-view unsupervised feature selection with adaptive similarity and view weight', ' two dimensional feature selection by sparse matrix regression', 'robust auto-weighted multi-view subspace clustering  with common subspace representation matrix' and 'scalable multi-view semi-supervised classification via adaptive regression' have been accepted by TKDE, TIP, Plos One and TIP respectively.  

Locations of Site Visitors

Updated on:2017-11-13 21:51      Total Visits:16451


Recently Visited

Similar Subject

Same institution

Similar Interests