Romadhoni, R.Wahyu (2022) ABSTRAK APLIKASI DETEKSI KEMIRIPAN WARNA MENGGUNAKAN METODE KNN(K-NEAREST NEIGHBORS). Skripsi thesis, UNIVERSITAS BHAYANGKARA SURABAYA.
![[img]](http://eprints.ubhara.ac.id/style/images/fileicons/text.png) |
Text
COVER DEPAN.pdf Download (322kB) |
|
Text
COVER.pdf Download (322kB) |
|
Text
LEMBAR PENGESAHAN.pdf Download (446kB) |
|
Text
LEMBAR PERSETUJUAN.pdf Download (361kB) |
|
Text
BAB 1.pdf Download (367kB) |
Abstract
Color can be defined as part of the addressing of the sense of sight, or as the nature of the light emitted. Human vision is so limited that determining or comparing the same colors will be difficult. A case in point is when going to buy wall paint. The eyes will find it difficult to choose the same paint color as the wall color. Data mining is a process of dredging or collecting important information from a large piece of data. The process of data mining uses statistical methods, mathematics, to utilize artificial intelligence technology. Classification is a technique in data mining to group data based on data attachment to sample data. One method that can be used is k-nn.while for clustering can use K-modes method. In this study, researchers used the image processing process by using the extraction of RGB color features that were changed in the HSV color feature then histogram. This feature can compare the accuracy of the level of color likeness with the KNN method. The criteria used are LAB Features and YcbCr Features. Based on testing using training data of 95 data and 10 color test data obtained application results have an average accuracy of 78% and errors of 22% for Kmodes while in KKN obtained an average accuracy value of 64% and error of 36%. K-modes cluster test obtained the highest accuracy rate obtained in the first test with an accuracy value of 86.7% with the number of k values equal to 3 and cluster number 3 while the lowest accuracy value was obtained in the 4th test of 66%. In the K-nn Test, the highest accuracy rate was obtained in the third and fourth tests with an accuracy value of 72% with the number of k values equal to the lowest 5 obtained at the value of k equal to 10 by 56%.
| Item Type: | Thesis (Skripsi) |
|---|---|
| Uncontrolled Keywords: | Color, Wall Paint, Data mining, K-nn, K-modes, LAB, YcbCr |
| Divisions: | Faculty of Engineering > Bachelor of Informations Technology |
| Depositing User: | Perpus Ubhara Surabaya |
| Date Deposited: | 28 Jun 2022 03:24 |
| Last Modified: | 28 Jun 2022 03:24 |
| URI: | http://eprints.ubhara.ac.id/id/eprint/1228 |
Actions (login required)
 |
View Item |
