Classification of Poor Households in West Sumatra Province using Decision Tree Algorithm C4.5

Abstrak

The significant and increasingly complex issue of poverty poses a considerable challenge to Indonesia's development, including in the province of West Sumatra, where the poverty rate was 5.92% in 2022. The government has initiated programs to address poverty by focusing on the criteria of impoverished households. Data on impoverished households can be obtained through the National Socio-Economic Survey (Susenas). One method that can classify impoverished households is the decision tree. A decision tree is a flowchart that resembles a tree. The C4.5 algorithm used in this research can handle both discrete and continuous types, manage variables with missing values, and prune decision tree branches. The analysis results indicate that the primary factor for impoverished households is the number of household members. The model created using the C4.5 algorithm was tested using a confusion matrix resulting in  an accuracy of 69.89%, sensitivity of 71.15% for classifying regular households, and specificity of 68.72% for classifying impoverished households.