This study aims to classify regencies / cities in Indonesia based on the level of expenditure in the snack food group by applying the K-Nearest Neighbor (K-NN) algorithm, and K-NN which is optimized using Multi Objective Particle Swarm Optimization (MOPSO). The classification results with K-NN show that the accuracy value is 90% which is classified as good, but the f1_score value of 75.00%, pre…