The inability to fulfill basic human needs is how poverty is defined. To address this issue, the indonesian goverment implements various social assistance programs, one of which is Kartu Indonesia Pintar (KIP), aimed at providing free education to children aged 7-18 who are economically disadvantaged. However, in the distribution of aid in the Pangkalan sesai sub-district, distributing officers often face challenges due to the high number of eligible recipients applying, complex data requierements, and limited time for the officers. Distributing this social assistance accurately is crusial. Therefore, this research aims to determine the accuracy value for the data of potential recipients of the Kartu Indonesia Pintar (KIP to enhance the data verification process’s outcomes. To tackle this issue, the research employs the K-Nearest Neighbor (K-NN) algoritm and also employs feature selection using Information Gain to reduce less influential attributes. The data used consists of 1998 records of KIP beneficiaries from the 2023 in excel format, with 33 attributes. After performing data cleaning an Information Gain-based feature selection, the dataset is reduced to 1675 records, with 5 selected attributes. The best classification result in this study is achieved with ratios of 7:3 and 8:2, and a value of k = 5, yielding the highest accuracy of 98,21%. The lowest accuracy is obtained using a ratio of 9:1 with the same k value when not using Information Gain, resulting in an accuracy of 89,82%.
Copyrights © 2023