ANINDITA, SABELLA DINNA (2018) Pemodelan Persentase Balita Stunting di Indonesia Menggunakan Regresi Nonparametrik Spline Truncated. Undergraduate thesis, Institut Teknologi Sepuluh Nopember.
Preview |
Text
06211645000003-Undergraduate_Theses.pdf - Accepted Version Download (1MB) | Preview |
Abstract
Masalah balita stunting menunjukkan adanya masalah gizi buruk dari kondisi ibu, masa janin, dan masa bayi serta masalah kesehatan lain baik yang mempengaruhi secara langsung maupun tidak langsung. Negara Indonesia menempati peringkat ke 5 dunia dengan jumlah anak pendek terbanyak pada tahun 2011. Hingga tahun 2017 lalu, WHO menetapkan Indonesia sebagai Negara dengan status gizi buruk. Hal ini disebabkan jumlah penderita stunting sekitar 35,6 persen (melebihi batas toleransi stunting maksimal 20 persen dari jumlah keseluruhan balita). Penelitian ini bertujuan mengetahui karakteristik sosial demografi balita stunting di Indonesia serta mendapatkan model terbaik dari balita stunting di Indonesia menggunakan metode regresi nonparametrik spline truncated. Metode ini digunakan karena pola hubungan antara variabel prediktor dengan variabel respon tidak mengikuti pola tertentu dan perilaku berubah-ubah di beberapa interval tertentu. Hasil yang diperoleh dari penelitian ini adalah rata-rata persentase balita stunting di Indonesia sebesar 27,286% dimana persentase balita stunting tertinggi adalah Sulawesi Barat dengan persentase sebesar 39,7%. Model terbaik yang didapat dengan regresi non parametrik spline truncated adalah menggunakan kombinasi knot 3,1,2,2,3. Dari lima variabel independen yang digunakan (persentase balita mendapat imunisasi lengkap, persentase balita mendapat asi selama 6bulan, persentase ibu hamil resiko KEK, persentase bayi lahir mendapat IMD dan persentase rumah tangga memiliki sanitasi layak), semuanya memberikan pengaruh yang signifikan terhadap model. Nilai koefisien determinasi yang dihasilkan dari model ini adalah sebesar 80,77%.
===============================================================================Stunting toddler problems indicate malnutrition problems from the mother's condition, fetal period, and infancy and other health problems that affect both directly and indirectly. Indonesia already in fifth rank in the world with the largest number of short children in 2011. Indonesia's position is only better than India, China, Nigeria and Pakistan. Last year 2017, WHO set Indonesia as a country with malnutrition status. This is due to the number of stunting patients about 35.6 percent (exceeding the maximum stunting tolerance limit at 20 percent of total number of toddler). This study was aims to find out the social characteristics of stunting toddler in Indonesia and get the best model from stunting toddler in Indonesia using spline truncated nonparametric regression method. This method was used because the relationship pattern between the predictor variable and the response variable does not follow a certain pattern and the behavior varies at certain intervals. The results obtained from this study that the average percentage of toddlers stunting in Indonesia is 27.286% which is the highest percentage of stunting toddlers at West Sulawesi with a percentage is 39.7%. The best model obtained with spline truncated nonparametric regression is using a combination of knots 3,1,2,2,3. From the five independent variables that used all have a significant effect on the model and the coefficient of determination that found from this model is 80.77%.
| Item Type: | Thesis (Undergraduate) |
|---|---|
| Uncontrolled Keywords: | GCV, Regresi Nonparametrik Spline, Stunting, Titik Knot |
| Subjects: | Q Science > QA Mathematics > QA278.2 Regression Analysis. Logistic regression |
| Divisions: | Faculty of Mathematics and Science > Statistics > 49201-(S1) Undergraduate Thesis |
| Depositing User: | Sabella Dinna Anindita |
| Date Deposited: | 28 Jun 2021 08:16 |
| Last Modified: | 28 Jun 2021 08:16 |
| URI: | http://repository.its.ac.id/id/eprint/56772 |
Actions (login required)
 |
View Item |