Tools
Tools
Puppala, Sai Kumar (2025) AI-Driven Agricultural Yield Optimization: Intergrating Synthetic Data and Multi-Crop Rotation Analysis for Enhanced Prediction Accuracy. Masters thesis, Dublin, National College of Ireland.
Preview |
PDF (Master of Science)
Download (930kB) | Preview |
Preview |
PDF (Configuration Manual)
Download (360kB) | Preview |
Abstract
Agricultural systems experience growing challenges from climate variability, lack of data, and optimisation complexities, for which sophisticated computational techniques for sustainable yield prediction and crop rotation strategies are needed. This research explores the potential of combined synthetic data generation and multi-crop rotation analysis to advance AI-based agricultural systems in various climatic conditions.
The study applies four-task methodology: Enhanced Conditional Generative Adversarial Networks (CGANs) for generating synthetic agricultural data, reinforcement learning based on Deep Q-Network (DQN) for the optimization of multi-crop rotations, bidirectional Long Short-Term Memory (LSTM) networks for climate-resilient yield prediction, and complete ensemble evaluation for integrating the system. The framework uses USDA CropNet dataset with systematic statistical verification as well as performance assessment.
Results indicate large improvements in all system modules: generation of synthetic data achieved 80.6% quality with 53.3% increase in geographical coverage, optimization of rotations achieved 82% success rate (134.3% improvement upon baseline), accuracy in the prediction of climatic outcomes improved by 21.7% beyond laid-down thresholds. The integrated ensemble system maintained 60% of the overall system performance, with all components of the research questions being answered effectively and achieving statistical significance (p < 0.05).
These findings furnish foundational evidence for AI-driven agricultural optimisation, demonstrating measurable enhancements in predictive accuracy and the efficacy of rotational strategies. The study offers quantifiable standards for future developments in agricultural AI as well as proof of computational frameworks for challenging issues in agricultural sustainability.
| Item Type: | Thesis (Masters) |
|---|---|
| Supervisors: | Name Email Niculescu, Hamilton UNSPECIFIED |
| Uncontrolled Keywords: | Agricultural AI; Synthetic Data Generation; Crop Rotation Optimization; Climate Prediction; Deep Learning; Reinforcement Learning |
| Subjects: | S Agriculture > S Agriculture (General) Q Science > QH Natural history > QH301 Biology > Methods of research. Technique. Experimental biology > Data processing. Bioinformatics > Artificial intelligence Q Science > Q Science (General) > Self-organizing systems. Conscious automata > Artificial intelligence S Agriculture > S Agriculture (General) > Farming Industry |
| Divisions: | School of Computing > Master of Science in Data Analytics |
| Depositing User: | Ciara O'Brien |
| Date Deposited: | 03 Jul 2026 09:22 |
| Last Modified: | 03 Jul 2026 09:22 |
| URI: | https://norma.ncirl.ie/id/eprint/9450 |
Actions (login required)
 |
View Item |