TY - GEN
T1 - Spectral regression
T2 - 7th IEEE International Conference on Data Mining, ICDM 2007
AU - Cai, Deng
AU - He, Xiaofei
AU - Han, Jiawei
PY - 2007
Y1 - 2007
N2 - Recently the problem of dimensionality reduction (or, subspace learning) has received a lot of interests in many fields of information processing, including data mining, information retrieval, and pattern recognition. Some popular methods include Principal Component Analysis (PCA), Linear Discriminant Analysis (LDA) and Locality Preserving Projection (LPP). However, a disadvantage of all these approaches is that the learned projective functions are linear combinations of all the original features, thus it is often difficult to interpret the results. In this paper, we propose a novel dimensionality reduction framework, called Unified Sparse Subspace Learning (USSL), for learning sparse projections. USSL casts the problem of learning the projective functions into a regression framework, which facilitates the use of different kinds of regularizers. By using a L1-norm regularizer (lasso), the sparse projections can be efficiently computed. Experimental results on real world classification and clustering problems demonstrate the effectiveness of our method.
AB - Recently the problem of dimensionality reduction (or, subspace learning) has received a lot of interests in many fields of information processing, including data mining, information retrieval, and pattern recognition. Some popular methods include Principal Component Analysis (PCA), Linear Discriminant Analysis (LDA) and Locality Preserving Projection (LPP). However, a disadvantage of all these approaches is that the learned projective functions are linear combinations of all the original features, thus it is often difficult to interpret the results. In this paper, we propose a novel dimensionality reduction framework, called Unified Sparse Subspace Learning (USSL), for learning sparse projections. USSL casts the problem of learning the projective functions into a regression framework, which facilitates the use of different kinds of regularizers. By using a L1-norm regularizer (lasso), the sparse projections can be efficiently computed. Experimental results on real world classification and clustering problems demonstrate the effectiveness of our method.
UR - http://www.scopus.com/inward/record.url?scp=49749131620&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=49749131620&partnerID=8YFLogxK
U2 - 10.1109/ICDM.2007.89
DO - 10.1109/ICDM.2007.89
M3 - Conference contribution
AN - SCOPUS:49749131620
SN - 0769530184
SN - 9780769530185
T3 - Proceedings - IEEE International Conference on Data Mining, ICDM
SP - 73
EP - 82
BT - Proceedings of the 7th IEEE International Conference on Data Mining, ICDM 2007
PB - Institute of Electrical and Electronics Engineers Inc.
Y2 - 28 October 2007 through 31 October 2007
ER -