Nugroho, Fikhi (2018) Implementasi Extreme Learning Machine Untuk Deteksi Dini Infeksi Menular Seks (IMS) Pada Puskesmas Dinoyo Kota Malang. Sarjana thesis, Universitas Brawijaya.
Abstract
Infeksi Menular Seksual (IMS) merupakan masalah kesehatan masyarakat yang cukup besar di dunia. Insiden kasus IMS pada banyak negara berkembang seperti kegagalan dalam mendiagnosis dan memberikan pengobatan pada stadium dini dapat menimbulkan komplikasi serius. Jawa Timur cukup mengkhawatirkan mendapat peringkat 2 untuk kasus AIDS (Acquired Immune Deficiency Syndrome) terbesar tahun 2011. Kota Malang menempati urutan lima besar dari jumlah kasus AIDS di Provinsi Jawa Timur. Jumlah kasus AIDS dan IMS harus ditangani dengan meningkatkan pelayanan dan pendidikan tentang IMS. Penggunaan sistem cerdas dapat mempermudah mendeteksi dini IMS secara komputasi. Parameter inputan yang diperlukan berupa 39 fitur yang terdiri dari jenis kelamin, 9 faktor risiko, dan 29 gejala. Proses analisis identifikasi gejala dini IMS menggunakan Extreme Learning Machine (ELM). Implementasi ELM tidak memerlukan rule IMS yang terkait dengan aturan melainkan membandingkan hasil penentuan keduanya. Jadi jika terjadi perubahan perhitungan maupun ketentuan identifikasi tidak mempengaruhi perhitungan ELM. Metode ELM digunakan untuk menentukan penyakit IMS menjadi sejumlah 17 kelas. Hasil terbaik dari tiga skenario pengujian tingkat akurasi antara hasil perhitungan ELM dengan hasil diagnosis pakar sebesar 36.36% untuk rasio 90:10, 50% untuk 100 hidden layer, dan 31.82% untuk weight range -1 sampai 0.
English Abstract
Sexually Transmitted Infections (STI) is a major public health problem in the world. Incidence of STI cases in many developing countries such as failure in diagnosing and provide treatment at an early stage can lead to serious complications. East Java is being ranked the second largest AIDS cases (Acquired Immune Deficiency Syndrome) in 2011, and should be quite worried. Malang city ranks the top five of the number of AIDS cases in East Java Province. The number of AIDS and STI cases should be addressed by improving services and education about STIs. The use of intelligent systems can make it easier to detect early IMS computing. The required input parameters consist of 39 features consisting of 2 sexes, 9 risk factors, and 29 symptoms. The process of identifying early identification of STI symptoms in this case will implement Extreme Learning Machine (ELM). The implementation of ELM itself does not require IMS rules related to the exact rules but rather compares the results of both determinations. Thus, if there is a change of calculation or identification provisions, it does not affect the calculation of ELM. The ELM method is used to determine STI disease to a number of 17 classes. The best results of the three test scenarios of accuracy between ELM calculations and expert diagnosis results were 36.36% for the 90:10 ratio, 50% for 100 hidden layers, and 31.82% for the weight range of -1 to 1.
Item Type: | Thesis (Sarjana) |
---|---|
Identification Number: | SKR/FTIK/2018/279/051801969 |
Uncontrolled Keywords: | Prediksi, Infeksi Menular Seks, Artificial Neural Networks, Extreme Learning Machine |
Subjects: | 000 Computer science, information and general works > 004 Computer science > 004.015 1 Finite mathematic |
Divisions: | Fakultas Ilmu Komputer > Teknik Informatika |
Depositing User: | Yusuf Dwi N. |
Date Deposited: | 28 Jun 2018 03:33 |
Last Modified: | 27 Oct 2021 05:47 |
URI: | http://repository.ub.ac.id/id/eprint/11707 |
Preview |
Text
BAB VI.pdf Download (674kB) | Preview |
Preview |
Text
BAB VII.pdf Download (601kB) | Preview |
Preview |
Text
Daftar Pustaka.pdf Download (483kB) | Preview |
Preview |
Text
Lampiran.pdf Download (1MB) | Preview |
Preview |
Text
Bagian Depan.pdf Download (1MB) | Preview |
Preview |
Text
BAB I.pdf Download (656kB) | Preview |
Preview |
Text
BAB II.pdf Download (1MB) | Preview |
Preview |
Text
BAB IV.pdf Download (1MB) | Preview |
Preview |
Text
BAB III.pdf Download (720kB) | Preview |
Preview |
Text
BAB V.pdf Download (1MB) | Preview |
Actions (login required)
![]() |
View Item |