This Author published in this journals
All Journal SPIRIT
Muhammad Ainun Nadjib
Teknik Informatika, Universitas Merdeka, Pasuruan, Indonesia

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

Found 1 Documents
Search

Implementasi Metode K-Nearest Neighbor dalam Mengklasifikasikan Kesegaran Ikan Kuro Menggunakan Citra Muslim Alamsyah; Muhammad Ainun Nadjib
SPIRIT Vol 14, No 2 (2022): Spirit
Publisher : STMIK YADIKA BANGIL

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.53567/spirit.v14i2.246

Abstract

Kurofish (Eleutheronematetradactylum) is a type of fish that spreads throughout Indonesian waters, with different local names in each region. On the east coast of Sumatra it is known by the name of sukain fish while on the north coast of Java it is known as kuro fish. Freshness of fish is one of the benchmarks for consumers in choosing quality or good fish for consumption, because fresh fish is rich in protein and nutrients. Fish is also known to contain omega 3 fatty acids which are beneficial for brain growth, as well as calcium, vitamin D and phosphorus which are good for bones. However, the nutritional content contained in the fish may not be optimal anymore if it is consumed in a condition that is not fresh. Not only that, consumption of fish that is not fresh which leads to rotten conditions can make someone poisoned.Fish freshness checks can be done through microbiological and chemical analysis, but this method is less effective because it requires a lot of manpower, is quite expensive, and takes longer. For traders, the level of freshness of fish is determined in the traditional way, namely by observing, holding and smelling the smell of fish, sometimes there is also something that escapes observation so that there are still fish that are not fresh.To reduce these problems, the authors apply the K-Nearest Neighbor method in classifying the freshness of fish using images based on the color of the fish. By using the Kuro Fish type, using Matlab tools and in the results of the study using the K-Nearest Neighbor method with 40 training data and producing an accuracy of 100% and 16 test data with 7 correct data resulting in poor accuracy, which is 43 ,75%