K-Nearest Neighbor (K-NN) algorithm with Euclidean and Manhattan in classification of student graduation

K-Nearest Neighbor (K-NN) algorithm is a classification that has been proven to solve various problems. Two approaches can be used in this are K-NN with Euclidean and Manhattan. The research aims apply the Manhattan classify accuracy of graduation. Student graduation determined by variables gender, major, number first-semester credits, second-semester third-semester grade point on first semester, second third age. These determine student graduation, timely or untimely. implementation carried out using Rapidminer software. results were obtained after testing 380 training data 163 data. best system was achieved at K=7 value 85.28%. two algorithmic did not affect results. Furthermore, addition K completely accuracy.