TY - GEN
T1 - Empirical study of the universum SVM learning for high-dimensional data
AU - Cherkassky, Vladimir
AU - Dai, Wuyang
PY - 2009
Y1 - 2009
N2 - Many applications of machine learning involve sparse high-dimensional data, where the number of input features is (much) larger than the number of data samples, d≫n. Predictive modeling of such data is very ill-posed and prone to overfitting. Several recent studies for modeling high-dimensional data employ new learning methodology called Learning through Contradictions or Universum Learning due to Vapnik (1998,2006). This method incorporates a priori knowledge about application data, in the form of additional Universum samples, into the learning process. This paper investigates generalization properties of the Universum-SVM and how they are related to characteristics of the data. We describe practical conditions for evaluating the effectiveness of Random Averaging Universum.
AB - Many applications of machine learning involve sparse high-dimensional data, where the number of input features is (much) larger than the number of data samples, d≫n. Predictive modeling of such data is very ill-posed and prone to overfitting. Several recent studies for modeling high-dimensional data employ new learning methodology called Learning through Contradictions or Universum Learning due to Vapnik (1998,2006). This method incorporates a priori knowledge about application data, in the form of additional Universum samples, into the learning process. This paper investigates generalization properties of the Universum-SVM and how they are related to characteristics of the data. We describe practical conditions for evaluating the effectiveness of Random Averaging Universum.
UR - http://www.scopus.com/inward/record.url?scp=70350608327&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=70350608327&partnerID=8YFLogxK
U2 - 10.1007/978-3-642-04274-4_96
DO - 10.1007/978-3-642-04274-4_96
M3 - Conference contribution
AN - SCOPUS:70350608327
SN - 3642042732
SN - 9783642042737
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 932
EP - 941
BT - Artificial Neural Networks - ICANN 2009 - 19th International Conference, Proceedings
T2 - 19th International Conference on Artificial Neural Networks, ICANN 2009
Y2 - 14 September 2009 through 17 September 2009
ER -