Predictive Analytics for Tea Prices: A Multi-Model Evaluation Framework

S. Anandhu, Aswathy Vijayan, Anil Kuruvila, A.R. Durga, Pratheesh, P. Gopinath, F. Thasnimol and V.S. Adarsh

Indian Journal of Agricultural Marketing, 2025, vol. 39, issue 1

Abstract: Tea price volatility presents various challenges to producers, traders, and policy makers involved in the production, and marketing of tea. Due to fluctuations in tea prices the decisions by farmers and traders to market their produce may often go wrong. This affects the welfare of farmers especially the Small Tea Growers (STGs). The correct anticipation of prices is a basis for strategic planning, risk management, and smooth supply chain arrangements. Hence this study attempts to forecast the tea prices using historic data using different types of models. In this study, the focus is on developing and comparing different models aimed at price forecasting for North Indian tea. Regression models employing Linear Regression, Gradient Boosting Regression (GBR), Support Vector Regression (SVR), Random Forest Regression (RFR); Time series models such as Autoregressive Integrated Moving Average (ARIMA) and Seasonal Autoregressive Integrated Moving Average (SARIMA) and Deep learning models involving Artificial Neural Networks (ANN), Recurrent Neural Networks (RNN), Long Short-Term Memory (LSTM) networks and the Prophet model was used for forecasting the tea prices. Hybrid models combining ARIMA with LSTM and GARCH (Generalised Autoregressive Conditional Heteroskedasticity) were considered as well, to take advantage of the linearity of ARIMA models and the nonlinearity of LSTM and GARCH models. Mean Absolute Error (MAE), Mean Absolute Percentage Error (MAPE), Mean Squared Error (MSE), Root Mean Squared Error (RMSE) and R2 were the evaluation metrics used to identify the best models. According to the results, the hybrid model, ARIMA-LSTM performed much better compared with other single models since both linear trends and nonlinear dynamics can be captured by tea prices. This study is relevant to the use of advanced and hybrid modelling techniques in forecasting agricultural commodity prices. The findings of this study have practical applications in the tea industry, and this technical intervention will help them to take better-informed decisions in marketing tea.

Keywords: Agribusiness (search for similar items in EconPapers)
Date: 2025
References: Add references at CitEc
Citations:

Downloads: (external link)
https://ageconsearch.umn.edu/record/400046/files/P ... r%20Tea%20Prices.pdf (application/pdf)

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:ags:injagm:400046

DOI: 10.22004/ag.econ.400046

Access Statistics for this article

More articles in Indian Journal of Agricultural Marketing from Indian Society of Agricultural Marketing
Bibliographic data for series maintained by AgEcon Search ().