Skripsi
OPTIMASI TIME-INVARIANT FUZZY TIME SERIES MENGGUNAKAN PARTICLE SWARM OPTIMIZATION UNTUK MEMPREDIKSI HARGA HASIL PERTANIAN
Agricultural commodity prices, such as rice and red chili, tend to fluctuate and are difficult to predict. This situation complicates decision-making for farmers when planning planting and harvesting periods, for traders in managing distribution strategies, and for governments in designing price stabilization policies. To address this issue, the Time-Invariant Fuzzy Time Series (TIFTS) method is used as it can model uncertain historical data. However, TIFTS still has limitations, particularly in determining fuzzy intervals, which significantly affect prediction accuracy. Therefore, this study implements the Particle Swarm Optimization (PSO) algorithm to optimize the parameters used in TIFTS. Testing was conducted on weekly price data of rice and red chili over a two-year period. The results show that the combination of TIFTS and PSO successfully reduced the MAPE from 9.90% to 5.24% for red chili and from 0.74% to 0.25% for rice. PSO parameters such as the number of particles, iterations, inertia weight, and acceleration coefficients (c1 and c2) were found to influence prediction results. This combined approach proves effective and can be applied as a decision support system in the agricultural sector.
| Inventory Code | Barcode | Call Number | Location | Status |
|---|---|---|---|---|
| 2507005303 | T182222 | T1822222025 | Central Library (References) | Available but not for loan - Not for Loan |