LASIFIKASI INDEKS PEMBANGUNAN MANUSIA (IPM) DENGAN PENDEKATAN K-NEARSET NEIGHBOR (K-NN)

Abstract

Human development index (HDI) is one of measuring instrument of achieving quality of life of one region even country. There are three basic components of the Human Development Index compilers: health dimension, knowledge dimension, and decent living dimension. To measure the health dimension, we use life expectancy at birth, knowledge dimension is used combination of indicator of old school expectation and mean of school length, and life dimension suitable for use indicator ability of people purchasing power to some basic requirement seen from mean of expense per customized capita. Data mining works to gather information from a large amount of data. Jobs that are closely related to data mining are prediction models, group analysis, association analysis, and anomaly detection. One of the classification methods contained in data mining and is often used and produces a fairly good accuracy is the K-Nearset Neighbor (k-NN) method. The absence of research on the classification or grouping of Human Development Index with
K-Nearset Neighbor (k-NN) method will be done by using k-NN method with k value of 1, 5, and 10. With the ultimate goal of comparing the accuracy of kaslifikasi between value k on the k-NN method. The result of classification of IPM by using k-NN method with k value of 5 and 10 obtained classification accuracy of 91.43% which is the best classification accuracy, with sensitivity of 100% and 83.33%.

Key words: HDI , Classification, K-Nearset Neighbor, Accuracy