Incremental learning algorithm of least squares support vector machines based on Renyi entropy
Guanhua Zhao, Min Hao · 2009
This paper introduces Renyi entropy and incremental learning algorithm into the financial distress prediction of enterprises and advances an incremental learning algorithm on least squares support vector machine based on Renyi entropy. By analysis and comparison of LS-SVM incremental learning algorithm based on Renyi entropy,traditional LS-SVM algorithm and standard SVM algorithm,LS-SVM incremental learning algorithm based on the Renyi entropy are evidently superior to the traditional LS-SVM algorithm and the standard SVM in terms of the number of training samples and computing time. This confirms the effectiveness and superiority of the introduction of Renyi entropy into the field of financial distress prediction.