Claim Missing Document
Check
Articles

Found 12 Documents
Search

EXAMINING THE SOYBEAN COMPETITIVENESS IN CENTRAL JAVA: A POLICY ANALYSIS MATRIX APPROACH Setiawan, Avi Budi; Antriyandarti, Ernoiz; Yusuf, Mochammad; Bowo, Prasetyo Ari; Wiyanti, Dian Tri
Agrisocionomics: Jurnal Sosial Ekonomi Pertanian Vol 8, No 3 (2024): November 2024
Publisher : Faculty of Animal and Agricultural Science, Diponegoro University

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.14710/agrisocionomics.v8i3.21806

Abstract

Soybean is one of the leading food crops in Indonesia, but its dependence on imports is very high. The productivity of soybean yields in Indonesia is also far below that of other soybean-producing countries. This study aims to analyze the competitiveness of soybean farming in Central Java Province. Policy Analysis Matrix (PAM) is used to analyze the soybean competitiveness. The results showed that the from PAM model shows that the PCR value is 0.37, which means that soybean farming is competitive in the current market. Furthermore, the DRC value is 0.30, indicating that soybean commodities have a comparative advantage or are competitive in the market if they are perfectly competitive and there are no distortions. Soybeans have an NPCO value of 0.960, farmers are paid 96% of what they should be paid. It appears that soybean farmers are relatively disadvantaged based on the NPCO value. Furthermore, the NPCI value is 0.98. The research results indicate that the NPCI has a value less than one. This implies the existence of consumer input protection policies in the form of subsidies. Thus, to meet domestic demand for soybeans, producing for domestic is better than importing from other countries. The and competitiveness analysis results show that soybean farming is profitable and competitive. The existence of competitive and comparative advantages indicates that soybean farming is still feasible to be cultivated domestically, so efforts are needed to increase efficiency to reduce dependence on imports.
Analisis Performa Algoritma Decision Tree, Naive Bayes, K-Nearest Neighbor untuk Klasifikasi Zona Daerah Risiko Covid-19 di Indonesia Ainurrohmah, Ainurohmah; Wiyanti, Dian Tri
Jurnal Teknologi Informasi dan Ilmu Komputer Vol 10 No 1: Februari 2023
Publisher : Fakultas Ilmu Komputer, Universitas Brawijaya

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.25126/jtiik.2023105935

Abstract

Pandemi Covid-19 terjadi di Indonesia. Pemerintah berupaya melakukan penanganan Covid-19, salah satunya dengan pembuatan peta risiko Covid-19. Peta risiko Covid-19 membagi zona berdasarkan Kabupaten/Kota. Zona risiko Covid-19 menjadi patokan pemerintah dalam mengambil kebijakan setiap daerah. Pemerintah menggunakan pembobotan dari 15 indikator untuk menentukan zona. Beberapa kali perubahan zona risiko Covid-19 pada website mengalami keterlambatan. Klasifikasi dapat menjadi alternatif penentuan zona risiko Covid-19, sehingga perubahan zona dapat dilakukan secara cepat dan efisien. Klasifikasi memiliki berbagai algoritma, setiap algoritma memiliki keunggulan dan kelemahan. Algoritma klasifikasi yang memiliki akurasi yang baik dengan waktu relatif cepat yaitu Decision Tree, Naïve Bayes dan K-Nearest Neighbor. Tujuan penelitian ini menghitung performa setiap algoritma, mendapatkan algoritma terbaik dan mendapatkan pola klasifikasi dari algoritma terbaik. Metode penelitian menggunakan 10-fold cross validation untuk pembagian data dan confusion matrix untuk menilai performa. Software yang digunakan yaitu Rapidminer dan WEKA. Hasil dari pengolahan data menunjukan semua algoritma mempunyai nilai performa yang baik yaitu diatas 70%. Semua algoritma tidak memerlukan waktu yang lama dalam pembuatan model. Nilai performa terbaik didapatkan dengan menggunakan algoritma decision tree dengan software WEKA dengan nilai performa 88% dan waktu 0,32 detik. Pola klasifikasi dari algoritma terbaik menghasilkan 77 aturan  yang membagi 3 zona klasifikasi yaitu rendah, sedang, dan tinggi. Atribut yang berpengaruh dalam klasifikasi zona risiko Covid-19 yaitu aktif, CR, CFR, laju insidensi, positif, dan meninggal. AbstractThe Covid-19 pandemic occurred in Indonesia. The government is trying to handle Covid-19, one of which is by making a Covid-19 risk map. The Covid-19 risk map divides zones based on Regency/City. The Covid-19 risk zone is the government's benchmark policy for each region. The government uses a weighting of 15 indicators to determine the zone. Several times the Covid-19 risk zone change on the website has been delayed. Classification can be an alternative to determining the Covid-19 risk zone,  that zone changes can be quickly and efficiently. Many algorithms can be used for classification. Several classification algorithms have good accuracy with relatively fast time are Decision Tree, K-Nearest Neighbor, and Naïve Bayes. The purpose of this study is to calculate the performance of each algorithm, get the best algorithm, and get the classification pattern from the best algorithm. The research method uses 10-fold cross validation for data sharing and confusion matrix to assess performance. The software used is Rapidminer. The results show that all algorithms have good performance values, which are above 70%. All algorithms do not require a long time in modeling. The best performance value using a Decision Tree algorithm. The classification pattern of the best algorithm produces 20 rules that divide 3 classification zones, namely low, medium, and high. Attributes that influence the classification of the Covid-19 risk zone are active, CR, CFR, incidence rate, positive, and death.