| NIM | K1B021065 |
|---|
| Namamhs | SELYA FAJRINA |
|---|
| Judul Artikel | Analisis Faktor-Faktor Signifikan yang Mempengaruhi Diabetes Melitus Menggunakan Forward Selection dan Best Subset Selection pada Regresi Logistik Biner |
|---|
| Abstrak (Bhs. Indonesia) | Kecamatan Jatilawang menempati urutan keempat kasus diabetes melitus tertinggi di Kabupaten Banyumas sejak 2021. Penelitian ini bertujuan mengidentifikasi faktor signifikan dan membangun model regresi logistik biner kejadian diabetes melitus. Data berasal dari riwayat pemeriksaan peserta program pengelolaan penyakit kronis di Puskesmas Jatilawang, dengan variabel prediktor: jenis kelamin, usia, hipertensi, indeks massa tubuh, dan aktivitas fisik. Pemilihan variabel terbaik menggunakan forward selection dan best subset selection, yang keduanya menghasilkan model sama, yaitu aktivitas fisik sebagai prediktor signifikan. Odds ratio 0,4117647 menunjukkan aktivitas fisik rendah menurunkan risiko diabetes melitus sebesar 58,83% dibanding aktivitas fisik tinggi. Model memiliki akurasi 77,23%, sensitivitas 100%, spesifisitas 77,23%, dan AUC 0,6003, sehingga layak digunakan meski kemampuan klasifikasinya belum optimal. |
|---|
| Abtrak (Bhs. Inggris) | Jatilawang subdistrict ranks fourth in the highest cases of diabetes mellitus in Banyumas regency since 2021. This study aims to identify significant factors and develop a binary logistic regression model for diabetes mellitus incidence. Data are sourced from the medical examination history of participants in the chronic disease management program at Jatilawang health center, with predictor variables: gender, age, hypertension, body mass index, and physical activity. The best variables were selected using forward selection and best subset selection, both of which produced the same model, with physical activity as a significant predictor. An odds ratio of 0.4117647 indicates that low physical activity reduces the risk of diabetes mellitus by 58.83% compared to high physical activity. The model has an accuracy of 77.23%, sensitivity of 100%, specificity of 77.23%, and AUC of 0.6003, thus it is suitable for use despite its classification ability not being optimal. |
|---|
| Kata kunci | Diabetes melitus, regresi logistik biner, forward selection, best subset selection, odds ratio. |
|---|
| Pembimbing 1 | Prof. Drs. Budi Pratikno, M.Stat.Sci., Ph.D. |
|---|
| Pembimbing 2 | Dr. Suroto, S.Si, M.Sc. |
|---|
| Pembimbing 3 | |
|---|
| Tahun | 2025 |
|---|
| Jumlah Halaman | 12 |
|---|
| Tgl. Entri | 2025-08-15 16:02:10.80493 |
|---|