SISTEM KLASIFIKASI PENENTUAN PENCAPAIAN TARGET BERDASARKAN DATA PRODUKSI PIPA MENGGUNAKAN KNN

Abstract

Data mining is a process
processing data to extract data in order to obtain data
certain patterns or information is more meaningful than a number of data
large in the database. One piece of data
mining is a classification used to group
object into a certain class based on the value of the attribute
related to the object being observed. The purpose of
This research is to build a classification system design
determining target achievement and conducting dataset testing
production on the K-Nearest Neighbor algorithm in
predict the determination of the achievement of production targets by
using predetermined criteria for
identify and classify production goods (pipes)
achieved or not achieved.
The use of the value of k can provide classification results
different values, the value of k that can give the result value
a good classification is the k-optimal value (the most k value
optimal). Then the k-fold in this study is k=3 with
the results of the accuracy rate of 88.58%. Evaluation in stages
to find the level of accuracy of the system by means of
distribution of training data and test data, testing is carried out with
one stage scenario, namely 80% training data: 20% test data
(80:20). The distribution of the dataset is 324 records divided into 260
for training data and 64 for test data from the total
data.