Year: 2024
Type: article
Abstract: The oilseed crops hold significant agricultural importance globally, particularly in India, where they contribute significantly to economic growth and food security.However, predicting its prices accurately remains challenging due to complex factors such as supply-demand dynamics, weather fluctuations, and global economic conditions.This study investigates price forecasting for major oilseed crops using traditional statistical models like autoregressive integrated moving average (ARIMA) and simple exponential smoothing (SES), alongside machine learning (ML) algorithms including artificial neural networks (ANN), random forest (RF), support vector regression (SVR), and K-nearest neighbors (KNN).The research employs a weighted ensemble (WE) approach based on the cuckoo search (CS) algorithm for weight optimization and model confidence set (MCS) for superior model selection.The ensemble model integrates predictions from multiple models, aiming to enhance forecast accuracy and reliability.Monthly wholesale price data of major oilseed crops from January 2010 to June 2024 across various markets from different states of India, sourced from AGMARKNET, have been analyzed.Results show that the proposed CS-WE model outperformed all the candidate models and suggest using an ensemble approach rather than a single model in managing risks and improving decision-making in the oilseed sector.
Authors: Ankit Kumar Singh, Ranjit Kumar Paul, Anita Sarkar, Md Yeasin, Kanchan Sinha, Soumen Pal, A.K Paul
Language: English
Field: Decision Sciences
Domain: Social Sciences
SDG: Zero hunger
Open Access status: bronze
OpenAlex ID: W4411653138