Performance Comparison of Different Classification Algorithms for Household Poverty Classification

There should be a way to identify poor households as the potential receiver of poverty alleviation programs if the aim of an organization is to serve the deprive one. This problem is addressed by using machine learning algorithms particularly the classification algorithms. In this study, data mining was used to analyze the poverty data that was extracted from the Community Based Monitoring System (CBMS) Database of Lagangilang, Abra, Philippines. Different methods of data mining classification were evaluated and compared like Naïve Bayes, ID3, Decision Tree, Logistic Regression and KNearest Neighbor (KNN). In order to get the best predictive model, the different classification algorithms were evaluated based on the performance metrics; accuracy, precision, recall, F1 score and AUC. The rapid miner was the datamining tool used to process the different classification algorithms and to analyze the dataset. The study showed that Naïve Bayes classifier is an efficient algorithm for predicting households that are poor and non-poor because it outperformed all the four algorithms in all the performance metrics used in the study. Furthermore, the error rate of the said algorithm is significant because it is 0.0014 only. Therefore, choosing the best classification algorithm for data mining task will give a better result.