<![CDATA[Harga beras rentan mengalami fluktuasi, berdampak signifikan pada kesejahteraan masyarakat, terutama kelompok berpendapatan rendah. Di Surabaya, kenaikan harga beras mendorong perlunya prediksi akurat untuk mitigasi dampak ekonomi. Penelitian ini bertujuan meramalkan harga beras menggunakan metode Fuzzy Time Series Cheng (FTS Cheng) yang dioptimalkan dengan Adaptive Particle Swarm Optimization (APSO) untuk menangani data non-linear dan fluktuatif. Data sekunder diambil dari Dinas Perindustrian dan Perdagangan Provinsi Jawa Timur (Siskaperbapo) periode 1 Januari 2023 hingga 31 Maret 2025, mencakup harga beras premium dan medium di Pasar Tambahrejo dan Pasar Wonokromo. Metode utama adalah FTS Cheng dengan optimasi APSO untuk meningkatkan akurasi prediksi. Model menunjukkan akurasi tinggi dengan MAPE (Mean Absolute Percentage Error) sangat rendah. Di Pasar Tambahrejo, MAPE beras premium 0,09% dan medium 0,00%. Di Pasar Wonokromo, MAPE premium 6,38% dan medium 0,85%. Optimasi APSO berhasil menurunkan MAPE, misalnya di Pasar Tambahrejo (premium turun 0,38%, medium turun 0,68%). Kombinasi FTS dan APSO menghasilkan prediksi harga beras yang presisi. Temuan ini dapat mendukung kebijakan stabilisasi harga, manajemen stok, dan perencanaan produksi beras lebih efektif, sekaligus meningkatkan stabilitas ekonomi rumah tangga. Abstract Rice prices are prone to fluctuations, significantly impacting public welfare, especially low-income groups. In Surabaya, rising rice prices necessitate accurate predictions to mitigate economic impacts. This research aims to forecast rice prices using the Fuzzy Time Series Cheng (FTS Cheng) method optimized with Adaptive Particle Swarm Optimization (APSO) to handle non-linear and fluctuating data. Secondary data was obtained from the East Java Provincial Department of Industry and Trade (Siskaperbapo) for the period January 1, 2023, to March 31, 2025, covering premium and medium rice prices at Tambahrejo Market and Wonokromo Market. The main method is FTS Cheng with APSO optimization to improve prediction accuracy. The model demonstrates high accuracy with very low MAPE (Mean Absolute Percentage Error). At Tambahrejo Market, MAPE for premium rice is 0.09% and medium rice is 0.00%. At Wonokromo Market, MAPE for premium rice is 6.38% and medium rice is 0.85%. APSO optimization successfully reduces MAPE, for example at Tambahrejo Market (premium decreased by 0.38%, medium decreased by 0.68%). The combination of FTS Cheng and APSO produces precise rice price predictions. These findings can support price stabilization policies, stock management, and more effective rice production planning, while improving household economic stability.]]>
Copyrights © 2026
