A Synergistic Approach to Forecasting Sesamum Prices in Andhra Pradesh: Hybrid, Machine Learning and Wavelet Decomposition Models
Shaik Shameem *
S.V. Agricultural College, Tirupati, Acharya N.G. Ranga Agricultural University, Andhra Pradesh, India.
Lavanya Kumari. P
Agricultural College, Bapatla, Acharya N.G. Ranga Agricultural University, Andhra Pradesh, India.
Ramana Murthy. B
S.V. Agricultural College, Tirupati, Acharya N.G. Ranga Agricultural University, Andhra Pradesh, India.
Vani. N
S.V. Agricultural College, Tirupati, Acharya N.G. Ranga Agricultural University, Andhra Pradesh, India.
*Author to whom correspondence should be addressed.
Abstract
Sesamum (Sesamum indicum L.) is an important oilseed crop in Andhra Pradesh, playing a vital role in India's agricultural economy through both domestic consumption and export. Despite its long history of cultivation, sesamum constitutes only a small fraction of global vegetable oil production. Its oil is highly valued for its nutritional quality, antioxidant properties, and stability, making it suitable for culinary, medicinal, and industrial uses. Accurate forecasting of wholesale sesamum prices is essential for stakeholders to make informed decisions and manage market risks efficiently. In this context, secondary data on sesamum prices in Andhra Pradesh from 2008 to 2024 was analysed using autocorrelation with the Box-Pierce test. A range of models were developed, including the ARIMA model and machine learning models such as ANN, SVR, ELM, as along with hybrid and wavelet-based models. Among these, the ARIMA+SVR hybrid model exhibited the highest predictive accuracy, thereby enhancing the reliability of price forecasts and contributing to improved planning, market stability, and economic efficiency in the agricultural sector.
Keywords: Sesamum prices, two-stage hybrid models, ARIMA, GARCH, wavelet decomposition models, machine learning models, RMSE, MSE, Andhra Pradesh