Puspitawuri, Annisa (2019) Diagnosis Tingkat Risiko Penyakit Stroke Menggunakan Metode K-Nearest Neighbor Dan Naïve Bayes. Sarjana thesis, Universitas Brawijaya.
Abstract
Penyakit stroke merupakan penyakit yang timbul akibat terputusnya suplai darah menuju otak karena terdapat semburan pembuluh darah atau terjadi sumbatan berupa gumpalan darah. Stroke merupakan penyebab kecacatan nomor satu dan penyebab kematian nomor tiga di dunia setelah penyakit jantung dan kanker, baik di negara maju maupun berkembang. Berdasarkan data Riset Kesehatan Dasar, prevalensi stroke di Indonesia pada tahun 2013 mengalami kenaikan jika dibandingkan dengan data Riskesdas 2007 dengan nilai angka 8,3%, naik mencapai angka 12,1% per 1.000 penduduk. Untuk itu diperlukan suatu tindakan pendeteksian tingkat risiko penyakit stroke agar dapat segera diatasi sesuai dengan tingkat risikonya. Penelitian ini mengusulkan adanya suatu aplikasi diagnosis tingkat risiko penyakit stroke menggunakan metode K-Nearest Neighbor dan Naïve Bayes karena data yang didapat menggunakan atribut numerik dan kategoris. Algoritme K-Nearest Neighbor digunakan untuk memproses data numerik, dan algoritme Naïve Bayes digunakan untuk memproses data kategoris. Hasil penelitian menunjukkan nilai akurasi tertinggi yang diperoleh pada data kelas seimbang adalah 96.67% dengan data latih 45, data uji 30 dan nilai K=15-22. Sedangkan pada data latih tidak seimbang, menunjukkan akurasi tertinggi sebesar 100% dengan jumlah data latih 60, data uji 30 dan nilai K=20-30
English Abstract
Stroke is a disease that arises due to the dissolution of blood supply to the brain because of bursts in the blood vessels or there was a blockage of blood clots. It is a number 3 cause of death after heart disease and cancer and stroke is a leading cause of disability, both in developed and developing countries. Based on Riskesdas data, stroke prevalence in Indonesia in 2013 has increased when compared with Riskesdas data in 2007 with a value of 8.3%, increase up to 12.1% per 1,000 population. Therefore, we need an action to detect the level of risk of stroke to be immediately addressed in accordance with the level of risk. This research proposes an application of diagnosis of stroke risk level using K-Nearest Neighbor and Naïve Bayes methods, because the data obtained using numerical and categorical attributes. K-Nearest Neighbor algorithm is used to process numerical data, and Naïve Bayes algorithm is used to process categorical data. The results showed that the highest accuracy value obtained in the balanced class data was 96.67% with 45 training datasets, 30 testing datasets and value of K=15-22. Meanwhile, the training datasets that is not balanced shows the highest accuracy of 100% with the number of training datasets is 60, 30 testing datasets and the value of K=20-30
Other obstract
-
Item Type: | Thesis (Sarjana) |
---|---|
Identification Number: | SKR/FILKOM/2019/75/051902245 |
Uncontrolled Keywords: | penyakit stroke, klasifikasi, K-Nearest Neighbor, Naïve Bayes, stroke, classification, K-Nearest Neighbor, Naïve Bayes |
Subjects: | 600 Technology (Applied sciences) > 616 Diseases > 616.8 Diseases of nervous system and mental disorders > 616.81 Cerebrovascular diseases |
Divisions: | Fakultas Ilmu Komputer > Teknik Informatika |
Depositing User: | Nur Cholis |
Date Deposited: | 23 Jun 2020 07:41 |
Last Modified: | 24 Oct 2021 03:25 |
URI: | http://repository.ub.ac.id/id/eprint/169292 |
Preview |
Text
Annisa Puspitawuri (2).pdf Download (5MB) | Preview |
Preview |
Text
Bagian Depan (2).pdf Download (1MB) | Preview |
Preview |
Text
Daftar Referensi (2).pdf Download (752kB) | Preview |
Actions (login required)
![]() |
View Item |