Article Per Year (5 Year)
p-Index From 2020 - 2025
0
P-Index
This Author published in this journals
All Journal Jurnal Pengembangan Teknologi Informasi dan Ilmu Komputer
Tri Halomoan Simanjuntak
Fakultas Ilmu Komputer, Universitas Brawijaya
Computer Science & IT Control & Systems Engineering Education Electrical & Electronics Engineering Engineering

Published : 1 Documents Claim Missing Document
Claim Missing Document
Check
Articles

Found 1 Documents
Search

 1 
Implementasi Modified K-Nearest Neighbor Dengan Otomatisasi Nilai K Pada Pengklasifikasian Penyakit Tanaman Kedelai Tri Halomoan Simanjuntak; Wayan Firdaus Mahmudy; Sutrisno Sutrisno
Jurnal Pengembangan Teknologi Informasi dan Ilmu Komputer Vol 1 No 2 (2017): Februari 2017
Publisher : Fakultas Ilmu Komputer (FILKOM), Universitas Brawijaya

Show Abstract | Download Original | Original Source | Check in Google Scholar | Full PDF (554.805 KB)

Abstract

Various diseases and pest attacks can cause serious problems to the soybean crop. One threat to the soybean crop development research centers and is the developer of the plant pests. Pests can reduce soybean yields by 80 % even if no serious control. Classification is needed to determine the types of diseases that attack soybean plants. This research use of Soybean Disease Data Set consisting of 266 training data and desktop-based applications to be built by implementing the algorithm Modified K - Nearest Neighbor, the parameter value of K is determined by the system using brute force methods to find the best K value. Each value of K with accuracy the best results will be recorded and used as the parameter value of K in the process of testing new data. K values in this method to define the number of nearest neighbors used for the classification process. The test results showed that the value of the parameter K affects the classification results and the accuracy result. Average accuracy tends to decrease with the addition of the value of k , while increasing the number of training data also accompanied by an increase in the accuracy of the results, for training data with imbalanced class accuracy values decreased with increasing amount of data. The results of the highest accuracy on the test at 100 % with a value of k = 1 and an average accuracy of 5 times the experimental is 98.83 %.
Co-Authors Sutrisno Sutrisno Wayan Firdaus Mahmudy