Forecasting India’s Spice Export Performance: A Comparative Analysis of Neural Networks and Fuzzy Time Series Models
A. Chandra Praphulla *
Department of Statistics and Computer Applications, ANGRAU, Bapatla, A.P., India.
K. Kiran Prakash
Department of Statistics and Computer Applications, ANGRAU, Bapatla, A.P., India.
D. Ramesh
Department of Statistics and Computer Applications, ANGRAU, Bapatla, A.P., India.
I.V.Y. Rama Rao
Agricultural Economics, ANGRAU, LAM, Guntur, A.P., India.
*Author to whom correspondence should be addressed.
Abstract
Agriculture contributes significantly to India’s economy, with exports playing a vital role in earning foreign exchange and supporting millions of farmers. India dominates global spice production and trade, being the top producer and exporter of Chilli, known for its pungency and quality, and Turmeric, renowned for its culinary and medicinal value. The purpose of this study is to determine the best model by comparing the Artificial Neural Network and Fuzzy Time Series. For this study, secondary data were collected from 1970-1971 to 2023-2024 and used to predict the Chilli and Turmeric exports in 2026-2027. The diagnostic criteria used in this study for comparison were Root Mean Square Error (RMSE), Mean Absolute Error (MAE), and Mean Absolute Percentage Error (MAPE). The results of this study showed that the Artificial Neural Network (ANN) method had the smallest error values compared to Fuzzy Time Series, which means it was more accurate in forecasting the export quantity of Chilli and Turmeric from India. In order to meet the increasing demand for Indian spices, it is important to enhance quality by improving farming practices.
Keywords: Chilli, Turmeric, export, ANN, fuzzy