Please use this identifier to cite or link to this item: http://repository.ipb.ac.id/handle/123456789/111226
Title: Identifikasi Faktor yang Memengaruhi Status Rawan Pangan Rumah Tangga di Provinsi Jawa Barat
Authors: Silvianti, Pika
Sartono, Bagus
Athallah, Syahla
Issue Date: 2022
Publisher: IPB University
Abstract: Kesejahteraan bagi suatu rumah tangga ditandai dengan terpenuhinya kebutuhan standar hidup salah satunya adalah pangan. Negara berkewajiban dalam memenuhi kebutuhan pangan tertuang dalam UU No.18 Tahun 2012. Perwujudan ketahanan pangan perlu dimulai dari terpenuhinya kebutuhan pangan pada skala terkecil (rumah tangga). Jawa Barat merupakan provinsi dengan jumlah penduduk terbanyak di Indonesia. Ketersediaan pangan akan terancam dengan semakin meningkatnya jumlah penduduk, dikhawatirkan dapat mengancam ketahanan pangan di Jawa Barat hingga ke skala rumah tangga. Penelitian ini membangun model klasifikasi status rawan pangan rumah tangga serta mengkaji faktor yang memengaruhinya dengan pendekatan metode multivariate adaptive regression splines (MARS). Model MARS terpilih pada penelitian ini terbentuk berdasarkan kombinasi basis function (BF) = 28, maximum interaction (MI) = 2, dan minimum observation (MO) = 1, model menghasilkan generalized cross validation (GCV) minimum sebesar 0.1604912. Terdapat 8 dari 14 peubah penjelas berpengaruh signifikan berdasarkan tingkat kepentingan terhadap model klasifikasi status rawan pangan rumah tangga di provinsi Jawa Barat. Evaluasi model dilakukan pada data uji dihasilkan nilai akurasi sebesar 64.66% dan apparent error rate (APER) sebesar 35.34%. Pada keseluruhan data dihasilkan nilai akurasi sebesar 64.9% dengan APER sebesar 35.1%. Statistik Uji Press’s Q menunjukan hasil klasifikasi sudah stabil dan konsisten secara statistik pada taraf nyata 5%.
Welfare for a household is characterized by the fulfilling living standard needs, one of which is food. The state is obliged to meet the food needs contained in Law No.18 of 2012. The realization of food security needs to start from the fulfilling food needs on the smallest scale (household). West Java is the province with the largest population in Indonesia. The increasing population will threaten food availability, feared to threaten food security in West Java to the scale of households. This study builds a classification model of household food insecurity status and examines the factors that influence it using multivariate adaptive regression splines (MARS) methods. The MARS model selected in the study was formed based on a combination of base functions (BF) = 28, maximum interaction (MI) = 2, and minimum observation (MO) = 1, the model produced a minimum generalized cross-validation (GCV) of 0.1604912. There are 8 out of 14 significant influences based on the level of importance to the classification model of household food insecurity status in West Java province. The evaluation of the model conducted on the test data resulted in an accuracy value of 64.66% and an apparent error rate (APER) of 35.34%. Overall data generated an accuracy value of 64.9% with an APER of 35.1%. Press's Q Test statistics show the classification results are stable and statistically consistent at a real level of 5%.
URI: http://repository.ipb.ac.id/handle/123456789/111226
Appears in Collections:UT - Statistics and Data Sciences

Files in This Item:
File Description SizeFormat 
Cover.pdf
  Restricted Access
Cover948.51 kBAdobe PDFView/Open
G14170040_Syahla Athallah.pdf
  Restricted Access
Full text1.26 MBAdobe PDFView/Open
Lampiran.pdf
  Restricted Access
Lampiran265.54 kBAdobe PDFView/Open


Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.