The Use of Brown's Double Exponential Smoothing Method to Predict Harvest Yields in Horticultural Crops

Abstract

Agriculture stands as a pivotal sub-sector within the economy of North Aceh. Among its primary commodities are horticultural crops, encompassing the cultivation of vegetables, fruits, medicinal plants, and ornamental flora. In endeavors to boost agricultural productivity and efficiency, the utilization of harvest prediction methodologies has grown increasingly indispensable. This study relies on historical harvest data spanning from 2017 to 2022 to forecast crops such as leafy greens, fruits, and medicinal plants. The selected plants for prediction include spinach, water spinach, cucumber, banana, durian, rambutan, ginger, lesser galangal, and turmeric. Data analysis employs Brown's double exponential smoothing method, selecting the α (alpha) parameter that minimizes the Mean Absolute Percentage Error (MAPE) for accurate forecasting. Spinach is anticipated to yield 1239.9508 quintals, with an α (alpha) parameter of 0.9 and a MAPE of 38.46%. Water spinach is forecasted to yield 2069.75 quintals, with an α (alpha) parameter of 0.5 and a MAPE of 18.14%. Cucumber is projected to yield 1023.22432 quintals, with an α (alpha) parameter of 0.4 and a MAPE of 31.51%. Consequently, the highest projected yield is for water spinach at 2069,75 quintals.