Triatmaja, Idham (2017) Sistem Diagnosis Penyakit Sapi Menggunakan Metode Neighbor Weighted K-Nearest Neighbor Berbasis Android. Sarjana thesis, Universitas Brawijaya.
Abstract
Salah satu masalah dalam beternak sapi adalah infeksi penyakit. Infeksi penyakit yang timbul pada setiap sapi dapat berbeda sehingga diperlukan ketelitian dalam melakukan diagnosis. Peternak sapi terkadang sulit dalam mengetahui penyakit sapi karena terbatasnya pengetahuan peternak terhadap penyakit sapi. Sulitnya mencari tenaga medis seperti dokter hewan menjadi masalah peternak karena tidak dapat secara cepat menangani sapi yang terkena penyakit. Ketika sapi sapi terinfeksi penyakit, ahli/pakar seperti dokter hewan sangatlah diperlukan dalam mengatasi hal tersebut. Berdasarkan penjelasan masalah tersebut maka akan dirancang sebuah sistem dengan menggabungkan antara obyek penelitian penyakit sapi dan metode NWK-NN. Judul penelitian tersebut adalah “Sistem Diagnosis Penyakit Sapi dengan Metode Neighbour Weighted K-Nearest Neighbour berbasis Android”. Terdapat 11 Jenis penyakit sapi meliputi Abses, Ascariasis, Bloat, Bovine Ephemeral Fever (BEF), Endometritis, Entritis, Mastitis, Omphalitis, Pneumonia, Rentensio, dan Scabies. Sistem diagnosis penyakit sapi ini memiliki proses utama yaitu proses menghitung jarak Euclidean, menentukan data sejumlah K ketetanggaan, menentukan nilai keanggotaan setiap data terhadap setiap kelas menentukan nilai keanggotaan setiap kelas dan menentukan nilai keanggotaan terbesar. Nilai variabel K 5, 10, 15, 20 dan 25 memiliki akurasi rata-rata sebesar 97,56% sementara untuk nilai K > 25 memiliki akurasi rata-rata yang semakin menurun. Akurasi rata-rata yang stabil adalah K dengan nilai 5, 10, 15, 20 dan 25 dengan akurasi 97,56%.
English Abstract
Problems in raising cattle is infection of disease. Infection of diseases that arise in each cow can be different so that required accuracy in making the diagnosis (Pambudi, 2010). Cattle ranchers are sometimes difficult to know cow disease due to the limited knowledge of breeders of cattle disease. The difficulty of seeking medical personnel such as veterinarians becomes a problem for farmers because they can not quickly handle cows affected by the disease. When cattle are infected with disease, experts / experts such as veterinarians are very necessary in overcoming it. Based on the explanation of the problem it will be designed an expert system by combining the object of cow disease research and the NWK-NN method. The title of the study was "System of Cow Disease Diagnosis with Neighbors Weighted K-Nearest Neighbors Method based on Android". There are 11 types of cow disease including Abscess, Ascariasis, Bloat, Bovine Ephemeral Fever (BEF), Endometritis, Entritis, Mastitis, Omphalitis, Pneumonia, Rentensio, and Scabies. This cow disease diagnosis expert system has the main process of calculating Euclidean distance, determining the data of a number of neighborhoods, determining the membership value of each data for each class determining the membership value of each class and determining the largest membership value. The variable values of K 5, 10, 15, 20 and 25 have an average accuracy of 97.56% while for the value of K> 25 has a decreasing average accuracy. The stable average accuracy is K with values 5, 10, 15, 20 and 25 with an accuracy of 97.56%.
| Item Type: | Thesis (Sarjana) |
|---|---|
| Identification Number: | SKR/FTIK/2017/709/051708221 |
| Uncontrolled Keywords: | Sistem Pakar, Neighbor Weighted K-Nearest Neighbor, Penyakit Sapi, Android |
| Subjects: | 000 Computer science, information and general works > 006 Special computer methods > 006.3 Artificial intelligence > 006.31 Machine learning > 006.312 Data mining |
| Divisions: | Fakultas Ilmu Komputer > Teknik Informatika |
| Depositing User: | Yusuf Dwi N. |
| Date Deposited: | 12 Oct 2017 03:47 |
| Last Modified: | 11 Nov 2020 07:56 |
| URI: | http://repository.ub.ac.id/id/eprint/3628 |
Actions (login required)
 |
View Item |