Pemodelan Kasus Tuberkulosis di Provinsi Jawa Timur Menggunakan Multivariate Adaptive Regression Splines

Khoirina, Annisa (2024) Pemodelan Kasus Tuberkulosis di Provinsi Jawa Timur Menggunakan Multivariate Adaptive Regression Splines. Other thesis, Institut Teknologi Sepuluh Nopember.

Text 5003201018-Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only until 1 October 2026. Download (5MB) | Request a copy

Abstract

Tuberkulosis adalah penyakit menular yang disebabkan oleh bakteri Mycobacterium tuberculosis dan merupakan masalah kesehatan global, termasuk di Indonesia. Indonesia menempati posisi kedua setelah India dalam jumlah kasus tuberkulosis tertinggi di dunia. Provinsi Jawa Timur memiliki kasus tuberkulosis tertinggi kedua di Indonesia setelah Jawa Barat. Pada tahun 2022, persentase penemuan kasus tuberkulosis di Jawa Timur mencapai 73,3 persen dengan jumlah kasus 78.799 kasus. Tingginya kasus tuberkulosis di Jawa Timur perlu dilakukan penelitian untuk mengidentifikasi faktor-faktor yang menyebabkan tingginya kasus tuberkulosis. Pada penelitian ini, pola hubungan antara variabel respons dan prediktor tidak diketahui serta terdiri dari lima variabel prediktor, sehingga data dikategorikan berdimensi tinggi. Oleh karena itu, metode regresi nonparametrik yang tepat digunakan adalah Multivariate Adaptive Regression Splines (MARS). Pemodelan MARS pada penelitian dilakukan menggunakan kombinasi Basis Function (BF) 10,15, dan 20, Maksimum Interaction (MI) yaitu 1,2, dan 3, dan Minimum Observation (MO) yaitu 0, 1, 2, dan 3 serta penentuan model MARS terbaiknya didasarkan pada nilai Generalized Cross Validation (GCV) yang paling minimum. Pada penelitian ini, didapatkan Model MARS terbaik dengan nilai GCV paling minimum yaitu 36,1103 dari hasil kombinasi BF = 20, M1 = 3, dan MI = 1. Variabel prediktor yang paling berpengaruh terhadap persentase penemuan kasus tuberkulosis secara berurutan adalah variabel persentase penduduk usia 15-64 tahun yang merokok, persentase penduduk yang ber-PHBS, kepadatan penduduk, dan persentase penduduk yang memiliki keluhan penyakit sebulan terakhir.
========================================================================================================================
Tuberculosis is an infectious disease caused by the bacterium Mycobacterium tuberculosis and is a global health problem, including in Indonesia. Indonesia is second only to India in the highest number of tuberculosis cases in the world. East Java province has the second highest tuberculosis cases in Indonesia after West Java. In 2022, the percentage of TB case finding in East Java reached 73.3 per cent with 78,799 cases. The high number of tuberculosis cases in East Java requires research to identify the factors that cause the high number of tuberculosis cases. In this study, the relationship pattern between the response variable and predictors is unknown and consists of five predictor variables, so the data is categorised as high-dimensional. Therefore, the appropriate nonparametric regression method used is Multivariate Adaptive Regression Splines (MARS). MARS modelling in this study was conducted using a combination of Basis Function (BF) 10, 15, and 20, Maximum Interaction (MI) of 1, 2, and 3, and Minimum Observation (MO) of 0, 1, 2, and 3 and the determination of the best MARS model is based on the minimum Generalized Cross Validation (GCV) value. In this study, the best MARS model was obtained with the minimum GCV value of 36.1103 from the combination of BF = 20, M1 = 3, and MI = 1. The predictor variables that have the most influence on the percentage of tuberculosis case finding in order are the variables of the percentage of the population aged 15-64 years who smoke, the percentage of the population with PHBS, population density, and the percentage of the population who have complaints of illness in the last month.

Item Type:	Thesis (Other)
Uncontrolled Keywords:	East Java, GCV, MARS, Nonparamteric Regression, Tuberculosis,GCV, Jawa Timur, MARS, Regresi nonparametrik, Tuberkulosis.
Subjects:	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:	Annisa Khoirina
Date Deposited:	09 Aug 2024 02:37
Last Modified:	09 Aug 2024 02:37
URI:	http://repository.its.ac.id/id/eprint/114929

Actions (login required)

View Item