期刊名称：Faculty of Computer Science and Information Technology
出版年度：2008
卷号：0
期号：0
语种：English
出版社：Faculty of Computer Science and Information Technology
摘要：Prediction of bankruptcy is a necessary action by an employer as a first step incase of bankruptcy or anticipation. Many methods can be done to predictbankruptcy. One method used is the data mining techniques and variousalgorithms have been developed and applied. Performance data miningalgorithms becomes a consideration in the selection of algorithms to predictbankruptcy. Therefore, this study will try to compare the performance of twodata mining algorithms are decision tree and Naive Bayes. The purpose of thisstudy is to compare the level of accuracy that is owned by a decision treealgorithm and the nave Bayes, in predicting corporate bankruptcies. From thiscomparison will be known to the percentage of accuracy and error of trainingdata and test data used. Based on the percentage of accuracy and errors will beknown to the performance of each algorithm. The number of data samples usedwere 33 companies consisting of 22 active companies and 11 corporatebankruptcies. Types of companies that are used are manufacturing companies,wholesale (wholesale), retail (retail) and services. By using the tool Weka(Waikato Environment for Knowledge Analysis) 3-4, grades will just than iscorrectly classified instances, incorrectly classified, kappa statistic, the meanabsolute error, root mean squared error, relative absolute error, root relativesquared error. From the comparison of these values can be concluded that thealgorithm has superior performance is the Naive Bayes algorithm to achieve100% accuracy.
