Improving Crop Yield Forecasting with Hybrid CNN-GRU and ARIMA-GRU Models

Sahul Hameed, Sameer Ahamed (2025) Improving Crop Yield Forecasting with Hybrid CNN-GRU and ARIMA-GRU Models. Masters thesis, Dublin, National College of Ireland.

Preview	PDF (Master of Science) Download (909kB) | Preview
Preview	PDF (Configuration Manual) Download (1MB) | Preview

Abstract

Accurate forecasting of crop yields is essential in proper agricultural planning and resource distribution, as well as food security. This study provides a detailed modeling framework in which standard regression methods are combined with deep learning architectures to predict crop yield using environmental, agronomic, and climatological factors. The analysis used a well-organized pipeline, which started with the preprocessing of data and exploratory data analysis, then feature engineering and model development. A wide range of models: Lasso Regression, K-Nearest Neighbors, Decision Tree, Deep Neural Networks (DNN), Long Short-Term Memory (LSTM), CNN-GRU, and a new hybrid model of ARIMA-GRU were trained and tested on the standard regression metrics: Mean Absolute Error (MAE), Mean Squared Error (MSE), and Coefficient of Determination (R2). The findings revealed a definite trend in the performance with the traditional models providing a baseline accuracy and deep learning models providing much better prediction ability. The ARIMA-GRU hybrid model performed best among all of them with an R2 score of 0.986 and the least MAE and MSE and was able to capture not only the linear trends but also the nonlinear trends in the data of crop yield. This work finds that hybrid modeling methods provide a promising way to tackle complex problems in agricultural forecasting and that integrating statistical inference with deep learning has a great potential to assist intelligent and data-driven agricultural systems.

Item Type:	Thesis (Masters)
Supervisors:	Name Email Kelly, John UNSPECIFIED
Subjects:	H Social Sciences > HD Industries. Land use. Labor > Specific Industries > Agriculture Industry S Agriculture > S Agriculture (General) > Farming Industry Q Science > Q Science (General) > Self-organizing systems. Conscious automata > Machine learning
Divisions:	School of Computing > Master of Science in Data Analytics
Depositing User:	Ciara O'Brien
Date Deposited:	03 Jul 2026 10:01
Last Modified:	03 Jul 2026 10:01
URI:	https://norma.ncirl.ie/id/eprint/9456

Actions (login required)

View Item