Optimasi Stabilitas Harga Telur Antar Wilayah di Jawa Timur dengan Metode BONSAI

Wulandari, Eva Novia (2026) Optimasi Stabilitas Harga Telur Antar Wilayah di Jawa Timur dengan Metode BONSAI. Other thesis, Institut Teknologi Sepuluh Nopember.

Text 5002221050-Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only Download (2MB) | Request a copy

Abstract

Jawa Timur merupakan produsen telur ayam ras terbesar di Indonesia, namun masih menghadapi disparitas harga yang signifikan antarwilayah, khususnya antara wilayah produsen seperti Kediri dan wilayah konsumen seperti Surabaya. Disparitas harga tersebut menunjukkan bahwa distribusi dan keseimbangan pasar telur antarwilayah belum berjalan secara optimal, sehingga berpotensi menimbulkan ketidakstabilan atau fluktuasi harga dari waktu ke waktu. Berbagai penelitian terdahulu telah mengkaji pergerakan dan volatilitas harga telur di Jawa Timur menggunakan pendekatan ekonometrika, namun umumnya hanya berfokus pada identifikasi faktor penyebab serta peramalan tren harga, tanpa menghasilkan kebijakan optimasi yang dapat digunakan untuk menstabilkan harga secara aktif. Oleh karena itu, penelitian ini bertujuan membangun model matematis struktur jaringan bagi masalah stabilitas harga telur antarwilayah di Jawa Timur, sekaligus menerapkan metode BONSAI (Bayesian Optimization of Network Systems under uncertAInty) untuk mencari kebijakan harga yang optimal dan robust terhadap ketidakpastian pasar. Penelitian ini menggunakan data sekunder harga telur dan harga jagung pipilan harian dari sembilan wilayah di Jawa Timur periode Januari 2023 hingga Mei 2026 yang bersumber dari SISKAPERBAPO, dilengkapi data produksi telur, jumlah penduduk, dan konsumsi telur per kapita. Sistem stabilitas harga dimodelkan sebagai struktur jaringan berbentuk Directed Acyclic Graph dengan empat fungsi komponen, yaitu supply, demand, harga keseimbangan, dan indeks stabilitas harga menggunakan Coefficient of Variation. Variabel desain berupa harga telur sembilan wilayah, sedangkan parameter ketidakpastian berupa harga jagung regional dan jumlah penduduk dengan 15 skenario kombinasi. Algoritma BONSAI diterapkan menggunakan Gaussian Process sebagai surrogate dan Thompson Sampling sebagai strategi akuisisi dalam optimasi robust dua tahap (max–min). Hasil penelitian menunjukkan algoritma BONSAI menurunkan nilai CV (Coefficient of Variation) disparitas harga dari 4,0975% menjadi 1,7034% setelah 149 iterasi optimasi, dengan solusi yang terbukti robust pada seluruh 15 skenario ketidakpastian serta rekomendasi harga yang bervariasi sesuai karakteristik tiap wilayah.
=========================================================================================================================================
East Java is the largest producer of commercial chicken eggs in Indonesia; however, it still faces significant interregional price disparities, particularly between producer regions such as Kediri and consumer regions such as Surabaya. These price disparities indicate that the distribution and market equilibrium of eggs among regions have not operated optimally, thereby potentially causing price instability or fluctuations over time. Various previous studies have examined egg price movements and volatility in East Java using econometric approaches, but they generally focus only on identifying causal factors and forecasting price trends, without producing optimization policies that can be used to actively stabilize prices. Therefore, this study aims to develop a mathematical function network model for the problem of interregional egg price stability in East Java, as well as to apply the BONSAI (Bayesian Optimization of Network Systems under uncertAInty) method to obtain price policies that are optimal and robust against market uncertainty. This study uses secondary data on daily egg prices and corn prices from nine regions in East Java covering the period from January 2023 to May 2026, obtained from SISKAPERBAPO, complemented by data on egg production, population, and per capita egg consumption. The price stability system is modeled as a function network in the form of a Directed Acyclic Graph consisting of four component functions, namely supply, demand, equilibrium price, and a price stability index measured using the Coefficient of Variation. The design variables are the egg prices in the nine regions, while the uncertainty parameters consist of regional corn prices and population levels under 15 combined scenarios. The BONSAI algorithm is implemented using a Gaussian Process as the surrogate model and Thompson Sampling as the acquisition strategy in a two-stage robust (max–min) optimization framework. The results show that the BONSAI algorithm reduced the CV (Coefficient of Variation) of price disparities from 4.0975% to 1.7034% after 149 optimization iterations, with a solution that was proven to be robust across all 15 uncertainty scenarios and price recommendations that vary according to the characteristics of each region.

Item Type:	Thesis (Other)
Uncontrolled Keywords:	Stabilitas Harga Telur, Disparitas Harga, Function Network, Bayesian optimization, BONSAI, Optimasi Robust, Egg Price Stability, Price Disparities, Function Network, Bayesian Optimization, BONSAI, Robust Optimization
Subjects:	Q Science > QA Mathematics > QA279.5 Bayesian statistical decision theory.
Divisions:	Faculty of Science and Data Analytics (SCIENTICS) > Mathematics > 44201-(S1) Undergraduate Thesis
Depositing User:	Eva Novia Wulandari
Date Deposited:	31 Jul 2026 03:34
Last Modified:	31 Jul 2026 03:34
URI:	http://repository.its.ac.id/id/eprint/140951

Actions (login required)

View Item