Evaluation of classification algorithms with solution to class imbalance problem on zakat distribution dataset

Class imbalance is one of the most serious and significant issues in data mining classification studies. In recent years, there has been a lot of interest in the problem of imbalance in a variety ofreal-world applications, such as fraud detection, medical diagnosis, and text classification. Similarly, the Zakat distribution UiTM Perlis dataset was utilized for the evaluation class imbalance dataset. The study aims to evaluate the performance ofresampling technique, Synthetic Over-Sampling Technique (SMOTE) and to identify the best classifiers for class imbalanced Zakat datasets by comparing the classifiers performance. A sampling-based approach is proposed in this study to solve the imbalance dataset of Zakat distribution in UiTM Perlis. The evaluation is based on various metrics, which include accuracy, precision, recall, F-measure, and the ROC area. The algorithms that are considered in the study include Logistic Regression, Decision Tree (C.4.5), and Random Forest. The results indicate that Random Forest consistently performs well across all evaluation metrics after SMOTE has been implemented. It attains the highest levels of accuracy, precision, recall, F-measure, and ROC area. In conclusion, this research highlights the effectiveness of SMOTE in addressing class imbalance in the distribution of Zakat.