Prakoso, Aldo Hani and Dahnial Syauqy, S.T., M.T., M.Sc. and Rakhmadhany Primananda, S.T., M.Kom. (2024) Sistem Klasifikasi Ikan Teri Asin Berformalin dan Tidak Berformalin Menggunakan Metode KNearest Neighbor Berbasis Weighted Euclidean Distance. Sarjana thesis, Universitas Brawijaya.
Abstract
Sebagai negara maritim tentu Indonesia memiliki produk perikanan yang melimpah. Ikan teri merupakan salah satu hasil laut Indonesia yang populer dan umum diolah sebagai produk ikan asin. Selain menambah cita rasa, teknik pengasinan pada ikan teri dapat memperpanjang masa simpan ikan. Sayangnya kebanyakan industri tradisional pengolah ikan teri asin lalai terhadap aspek sanitasi dan higiene, sehingga ikan teri asin yang dihasilkan rentan terhadap bakteri penyebab pembusukan. Dalam upaya mengatasi masalah tersebut, beberapa produsen bahkan memanfaatkan bahan kimia berbahaya seperti formalin yang dapat memicu penyakit kanker. Untuk melindungi konsumen dari bahaya formalin, penelitian ini mengembangkan sistem pengklasifikasi ikan teri asin berformalin dan tidak berformalin berdasarkan fitur HCHO dan fitur warna. Sebelumnya telah ada penelitian terkait sistem serupa dengan objek klasifikasi berbeda, sayangnya dalam penelitian tersebut metode klasifikasi knearest neighbor yang digunakan memiliki kelemahan berupa ketimpangan antara fitur warna dan fitur HCHO yang menyebabkan fitur HCHO memiliki pengaruh yang sangat kecil dalam proses klasifikasi. Berangkat dari permasalahan tersebut, pada penelitian kali ini akan digunakan metode knearest neighbor berbasis weighted euclidean distance dan penerapan normalisasi fitur guna menyeimbangkan pengaruh fitur warna dan fitur HCHO. Berdasarkan hasil pengujian, metode knearest neighbor berbasis weighted euclidean distance dengan penerapan normalisasi mampu mencatatkan nilai akurasi mencapai 100%, meningkat sebesar 12.01% dibanding akurasi metode knearest neighbor pada penelitian terdahulu yang mencatatkan nilai akurasi sebesar 87.99% ketika diterapkan dan diuji pada sistem di penelitian ini.
English Abstract
As a maritime country, Indonesia has abundant fishery products. Anchovy is one of Indonesia's popular marine products and is commonly processed as a salted fish product. In addition to adding flavor, salting techniques on anchovies can extend the shelf life of the fish. Unfortunately, most traditional processing industries of salted anchovies neglect sanitation and hygiene aspects, so the resulting salted anchovies are susceptible to bacteria that cause spoilage. In an effort to overcome this problem, some producers even utilize harmful chemicals such as formalin, which can trigger cancer. To protect consumers from the dangers of formalin, this research developed a system for classifying formalined and unformalined salted anchovies based on HCHO features and color features. Previously there has been research related to similar systems with different classification objects, unfortunately in that research the knearest neighbor classification method used has a weakness in the form of an imbalance between color features and HCHO features which causes the HCHO feature have very little influence in the classification process. Departing from these problems, this research will use the knearest neighbor method based on weighted euclidean distance and the application of feature normalization to balance the influence of color features and HCHO features. Based on the test results, the knearest neighbor method based on weighted euclidean distance and the application of feature normalization was able to record an accuracy value of 100%, an increase of 12.01% compared to the accuracy of the knearest neighbor method in previous research which recorded an accuracy value of 87.99% when applied and tested on the system in this research.
| Item Type: | Thesis (Sarjana) |
|---|---|
| Identification Number: | 0524150011 |
| Uncontrolled Keywords: | teri asin, formalin, knearest neighbor, weighted euclidean distance, normalisasi-salted anchovies, formalin, knearest neighbor, weighted euclidean distance, normalization |
| Divisions: | Fakultas Ilmu Komputer > Teknik Komputer |
| Depositing User: | Sugeng Moelyono |
| Date Deposited: | 13 Feb 2024 04:28 |
| Last Modified: | 13 Feb 2024 04:28 |
| URI: | http://repository.ub.ac.id/id/eprint/214239 |
![[thumbnail of DALAM MASA EMBARGO]](http://repository.ub.ac.id/style/images/fileicons/text.png) |
Text (DALAM MASA EMBARGO)
ALDO HANI PRAKOSO.pdf Restricted to Registered users only Download (2MB) |
Actions (login required)
 |
View Item |