Application of SMOTE technique for imbalanced breast cancer data set

Categorization data sets that have uneven class compositions are said to be unbalanced datasets. When a decision system is built to find a rare but significant occurrence, there are often substantial imbalances in real- world domains. The synthetic minority oversampling approach, or SMOTE, is one of the oversampling techniques most frequently employed to address the imbalance issue. By duplicating and adding more minority class samples at random, it seeks to balance the distribution of classes. SMOTE produces fresh minority instances by combining minority examples that already exist. Successful minority case prediction may be hampered by an unbalanced class distribution. This might have a significant impact on industries like medicine. Positive examples that are rare but substantial may be misclassified due to the overwhelming prevalence of the dominant class (negative cases). The data was collected from the UCI website which is Breast Cancer Data set. The algorithm chosen is J48, Random Forest and Logistic based on a little research about the best algorithm in another article The aim of this study is to apply SMOTE technique to an imbalanced breast cancer data set. By using WEKA as a tool with a different percentage of imbalance ratio to get the accurate output. Then, to validate the performance of those algorithms chosen by using Friedman Test with Rommel's Post-Hoc procedure analysis.