论文信息 - Novel SVM Performance Estimators for Information Retrieval Systems

Novel SVM Performance Estimators for Information Retrieval Systems

SVMs (support vector machines) have met with a significant success in information retrieval field, especially handling text classification tasks. Although various performance estimators for SVMs have been proposed, they only focus on the accuracy based on the LOO (leave-one-out) cross validation procedure. The information-retrieval-related performance measures are always neglected in kernel learning methodology. In this paper, we have proposed a set of information-retrieval-oriented performance estimators for SVMs, which are based on the span bound of the LOO procedure. Experiments have proved that our proposed estimators are both effective and stable. ...

Hui Song | Shui Yu | Fanyuan Ma

[1] V. Vapnik,et al. Bounds on Error Expectation for Support Vector Machines , 2000, Neural Computation.

[2] Sayan Mukherjee,et al. Choosing Multiple Parameters for Support Vector Machines , 2002, Machine Learning.

[3] Sebastian Thrun,et al. Learning to Classify Text from Labeled and Unlabeled Documents , 1998, AAAI/IAAI.

[4] Bernhard Schölkopf,et al. GACV for Support Vector Machines , 2000 .

[5] Thorsten Joachims,et al. Estimating the Generalization Performance of an SVM Efficiently , 2000, ICML.

[6] Thorsten Joachims,et al. Transductive Inference for Text Classification using Support Vector Machines , 1999, ICML.

[7] Thorsten Joachims,et al. Text Categorization with Support Vector Machines: Learning with Many Relevant Features , 1998, ECML.