Crop Yeild prediction using ML and Quantum computing

Crop yield prediction is a critical challenge in precision agriculture, requiring accurate analysis of soil and environmental parameters to support informed decision-making. This research presents a hybrid crop yield prediction framework that integrates classical machine learning with quantum-inspired computing techniques to enhance prediction accuracy and robustness. The proposed system utilizes key agronomic features, including Nitrogen (N), Phosphorus (P), Potassium (K), temperature, humidity, soil pH, and rainfall, collected through user input and preprocessed using feature scaling.

To capture complex nonlinear relationships within agricultural data, quantum feature encoding is employed by transforming classical inputs into entangled quantum state representations. These quantum-encoded features are evaluated using multiple kernel-based learning algorithms, namely Quantum Kernel Support Vector Machine (QKSVM), Quantum Kernel Logistic Regression (QKLR), and Quantum Kernel Random Forest operating in the kernel space. Among the evaluated models, the Quantum Kernel SVM demonstrated superior performance, achieving an accuracy of 0.97, significantly outperforming the remaining quantum kernel classifiers. This improvement highlights the effectiveness of quantum kernel similarity measures in high-dimensional agricultural datasets.

The proposed system is implemented using the Django web framework, enabling secure user authentication and real-time crop prediction. In addition to identifying the most suitable crop under given conditions, the framework provides yield interpretation and crop-specific agronomic recommendations to assist farmers in optimizing cultivation strategies. The experimental results confirm that combining machine learning with quantum-inspired techniques offers a promising direction for intelligent, data-driven, and sustainable agricultural systems.

Keywords

Crop Yield Prediction, Quantum Machine Learning, Quantum Kernel SVM, Precision Agriculture, Django Framework, Agronomic Decision Support, Hybrid Learning Models