Melis, Gusde (2019) Propensity Score Matching (PSM) Menggunakan Multivariate Adaptive Regresi Spline (MARS) pada Kasus HIV/AIDS. Other thesis, Institut Teknologi Sepuluh Nopember.
Preview |
Text
06211745000004_Undergraduate_Theses.pdf - Published Version Download (2MB) | Preview |
Abstract
Permasalahan HIV/AIDS menjadi tantangan kesehatan hampir di seluruh dunia, termasuk Indonesia. Kementerian Kesehatan melaporkan bahwa angka penderita HIV terus meningkat setiap tahun. Jumlah kumulatif infeksi HIV yang dilaporkan sampai Desember 2017 adalah 280.623 jiwa meningkat menjadi 301.959 jiwa per Juni 2018. Akibat paling fatal dari HIV/AIDS adalah kematian yang lebih dari 90% disebabkan oleh Infeksi Oportunistik (IO). Sebagai upaya untuk mengurangi kematian penderita HIV/AIDS perlu diketahui faktor-faktor yang mempengaruhi terjadinya IO. Namun, karena terdapat kemungkinan variabel confounding pada kejadian IO penderita HIV/AIDS yang mengakibatkan hasil estimasi menjadi bias dan menjadi tidak akurat. Sehingga, dilakukan penelitian yang bertujuan untuk mengurangi efek bias yang disebabkan oleh variabel confounding. Metode yang digunakan adalah Propensity Score Matching (PSM) dengan metode estimasi Multivariate Adaptive Regresi Spline (MARS). Hasil penelitian menunjukkan bahwa pemberian terapi ARV merupakan variabel yang terpilih sebagai confounding dan model terbaik MARS diperoleh menggunakan BF=12, MI=2, dan MO=2 dengan ketepatan klasifikasi model 76%. Hasil estimasi ATT menunjukkan bahwa pemberian terapi ARV berpengaruh signifikan terhadap IO. Nilai bias yang dapat direduksi PSM menggunakan MARS adalah 37,99%.
==============================================================================================================================
HIV/AIDS became a problem for health almost all over the world includes Indonesia. The Ministry of Health reported that the suffers of HIV have increased continuously every year. The cumulative number of HIV infection since December 2017 was 280.623 soul increasing to 307.959 soul as June 2018. The most consequence of HIV/AIDS is deaths of more than 90% due to Oportunistic Infection (OI). As the effort for reducing the death of people with HIV/AIDS, it is necessary to know the factors that influence the occurrence of OI. However, there is possibility of confounding variable on the incidence of OI with HIV/AIDS sufferer that cause the result of estimate are biased and inaccurate. Therefore, research was conducted to reduce bias effect caused by confounding variable. The method used Propensity Score Matching (PSM) with an estimated method of Multivariate Adaptive Regresi Spline (MARS). The results showed that the provision of ARV therapy was the variable chosen as confounding variable and the best MARS model was obtained using BF=12, MI=2, and MO=2 with the classification accuracy of the MARS model was 76%. The estimation of ATT showed the the provision of ARV therapy was significant to OI. The bias value that could be reduced by PSM with MARS was 37,99%.
| Item Type: | Thesis (Other) |
|---|---|
| Additional Information: | RSSt 519.535 Mel p-1 2019 |
| Uncontrolled Keywords: | HIV/AIDS, IO, MARS, PSM |
| Subjects: | Q Science Q Science > QA Mathematics Q Science > QA Mathematics > QA278.2 Regression Analysis. Logistic regression Q Science > QA Mathematics > QA76.6 Computer programming. Q Science > QA Mathematics > QA278 Cluster Analysis. Multivariate analysis. Correspondence analysis (Statistics) |
| Divisions: | Faculty of Science and Data Analytics (SCIENTICS) > Statistics > 49201-(S1) Undergraduate Thesis |
| Depositing User: | Gusde Melis |
| Date Deposited: | 27 Dec 2022 01:23 |
| Last Modified: | 27 Dec 2022 01:23 |
| URI: | http://repository.its.ac.id/id/eprint/63390 |
Actions (login required)
 |
View Item |