<![CDATA[The competence in predicting financial distress becomes an important research due tothe advantage in preventing companies financial failure. Besides, financial distressprediction model will give benefit to the investors and creditors. This research developa financial distress prediction model for listed manufacturing companies in Indonesiausing Support Vector Machines (SVM). Mathematically, SVM is formulated in the formof quadratic programming, which requires high computational time in finding theoptimal solution. In this research, Cross Entropy (CE) is used to optimize one of theSVM’s parameter that is Lagrange multipliers to find the optimal solution or nearoptimal solution of dual Lagrange SVM. The accuracy of the prediction model andcomputation time will be compared between standard SVM and CE-SVM. Finally, notethat the CE-SVM can solve classification problems in computing time 9.7 times shorterthan the standard SVM with good accuracy results. Keywords: cross entropy, lagrange multipliers, support vector machines, financialdistress]]>
Copyrights © 2016
