Initial Seed Value Efficiency on Data Mining Tools Performances: A Credit Approval Classification Example

After 2000s, Computer capacities and features are increased access to data made easy. However, the produced recorded should be meaningful. Transformation of unprocessed into meaningful information can done with help mining. In this study, classification methods from mining applications studied. First, parameters that make results same set different were investigated on 4 tools (Weka, Rapid Miner, Knime, Orange), It has been tested 3 algorithms (K nearest neighborhood, Naive Bayes, Random Forest). order evaluate performance while creating models, was divided training test as 80% -20%, 70% -30% 60-40%. The accuracy, roc precision values used classifying data. While classifying, effect algorithm is observed. most important these initial seed value. a value using especially in determines placement directly affects result. respect, it very determine correctly. between 0 100 evaluated shown could change accuracy approximately by 5%.