Yoshep Paulus Apri Caraka Yuda
BPS-Statistics Indonesia

Published : 2 Documents Claim Missing Document
Claim Missing Document
Check
Articles

Found 2 Documents
Search

Identifikasi Data Outlier (Pencilan) dan Kenormalan Data Pada Data Univariat serta Alternatif Penyelesaiannya Pardomuan Robinson Sihombing; Suryadiningrat Suryadiningrat; Deden Achmad Sunarjo; Yoshep Paulus Apri Caraka Yuda
Jurnal Ekonomi Dan Statistik Indonesia Vol 2 No 3 (2022): Berdikari: Jurnal Ekonomi dan Statistik Indonesia (JESI)
Publisher : Future Science

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.11594/jesi.02.03.07

Abstract

Penelitian ini bertujuan mengindentifikasi outlier (pencilan) dan kenormalan data pada univariat data. Adapun data yang digunakan berupa data persentase kemiskinan di Indonesia tahun 2022 yang berasal dari Badan Pusat Statistik. Metode pengujian outlier dilakukan dengan menggunakan grafik box plot, histrogram dan uji Grubbs. Sedangkan pengujian kenormalan data menggunkan uji SK Test dan Shapiro Wilk. Hasil penelitian menunjukkan terdapat data outlier yaitu pada observasi Provinsi Papua, dan data tidak berdistribusi normal. Selanjutnya dilakukan berbagai alternatif dalam menangani data outlier. Hasil menunjukkan menggunakan teknik tranformasi box cox, winsorizing dan trimming data, dapat menyelesaikan masalah outlier data. Metode box cox dan trimming sekaligus mampu mengatasi masalah kenormalan data, sedangkan metode winsorizing belum dapat mengatasi masalah kenormalan data.
KOMPARASI PERFORMA K-MEANS DAN FUZZY C-MEANS: Studi Kasus: Indeks Pembangunan Ekonomi Inkslusif Indonesia Pardomuan Robinson Sihombing; Yoshep Paulus Apri Caraka Yuda; Busminoloan Busminoloan; Iis Hayyun Nurul Islam
Jurnal Bayesian : Jurnal Ilmiah Statistika dan Ekonometrika Vol. 2 No. 2 (2022): Jurnal Bayesian : Jurnal Ilmiah Statistika dan Ekonometrika
Publisher : LPPM Universitas Bina Bangsa

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.46306/bay.v2i2.35

Abstract

This study aims to test the performance of the K-Means Cluster method with Fuzzy C-Means. The data used is data from the Inclusive Economic Development Index in 34 provinces in Indonesia in 2021. The data is sourced from Bappenas. The optimum number of clusters suggested using the Elbow method technique is as many as 4 clusters. By paying attention to the silouhette value the K-Means method is as good as the Fuzzi C-Means. However, the K-Means method is better than the Fuzzy C-Means model when viewed based on the criteria of smaller AIC and BIC values and a larger R 2. The provinces of Papua and West Papua have negative cluster means values for all variables so it is said that it is still lacking for all pillars of the IEDI. On the other hand, the provinces of DI Yogyakarta and DKI Jakarta have positive cluster means values for all variables so that they are said to be good in terms of the economy and opportunities and access but still have high inequality and poverty. Comprehensive and targeted policies are needed so that inclusive economic development in Indonesia can be evenly distributed and increased every year