论文信息 - Multi-Class Prediction of Weighted Support Vector Machine Based on Temporal Data

Multi-Class Prediction of Weighted Support Vector Machine Based on Temporal Data

In this paper, we discuss a problem of multi-classification on temporal data. First, we present a classification model. According to the characteristic of temporal data and the need of the fact, we propose a weighted support vector multi-classification algorithm and analyze its classification performance theoretically. This algorithm introduces weight factors for samples and classes. Finally, experiments show that our proposed algorithm can effectively solve the misclassification problems and improves the correctness of multi-classification

Lifang Peng | Gengui Zhou | Zhiqing Meng | Yihua Zhu

[1] Sun De-shan. Application of LS-SVM to Prediction of Chaotic Time Series , 2004 .

[2] Yang Wei. Discovery of Association Rules with Temporal Constraint in Databases , 1999 .

[3] He Wei. Study on the Support Vector Machines Model for Sales Volume Prediction and Parameters Selection , 2005 .

[4] Tang Chang. Mine the Quasi Periodicity from Web Data , 2000 .

[5] Vladimir N. Vapnik,et al. The Nature of Statistical Learning Theory , 2000, Statistics for Engineering and Information Science.

[6] Francis Eng Hock Tay,et al. Modified support vector machines in financial time series forecasting , 2002, Neurocomputing.

[7] Alexander J. Smola,et al. Learning with kernels , 1998 .

[8] Sushil Jajodia,et al. Logical design for temporal databases with multiple granularities , 1997, TODS.

[9] Lijuan Cao,et al. Dynamic support vector machines for non-stationary time series forecasting , 2002, Intell. Data Anal..