Comparison of the Effects of Cross-validation Methods on Determining Performances of Classifiers Used in Diagnosing Congestive Heart Failure