Perbandingan Kinerja Pengklasifikasi Citra Buah Kakao Sakit dan Sehat Menggunakan Support Vector Machine dan K-Nearest Neighbors

BLIKON, Yohanes Balawuri (2023) Perbandingan Kinerja Pengklasifikasi Citra Buah Kakao Sakit dan Sehat Menggunakan Support Vector Machine dan K-Nearest Neighbors. Undergraduate thesis, Universitas Katolik Widya Mandira Kupang.

	Text ABSTRAK.pdf Download (549kB)
	Text BAB I.pdf Download (194kB)
	Text BAB II.pdf Restricted to Repository staff only Download (233kB)
	Text BAB III.pdf Restricted to Repository staff only Download (196kB)
	Text BAB IV.pdf Restricted to Repository staff only Download (466kB)
	Text BAB V.pdf Restricted to Repository staff only Download (194kB)
	Text BAB VI.pdf Download (746kB)

Abstract

Cocoa is one of the crops in the plantation sector. Cocoa plantations withthe result that cocoa beans can be processed into the basic ingredients of flour orchocolate. The existence of these plantations certainly needs the support of technology or artificial intelligence by building a Support Vector Machine (SVM) and K-Nearest Neighbors (KNN) classification model to be applied totheindustrial sector through various adjustments. This study compares the SVM and KNN classification models on a dataset of sick and healthy cocoa pods of 4,390 images and 3 classes with the aim of knowing a more precise classifier performance. From the results of the trials conducted, the performance of the SVM classification model with the Radial Basis Function (RBF) kernel type andcross validation 2 obtained a higher prediction result of 82.5%, while the KNN classification model with number of neighbors 5, metric euclidean, and weight uniform accuracy rate of 82.4%.

Actions (login required)

View Item