Communications in computer and information science

Quantum Reinforcement Learning Framework for Agricultural Seed Treatment Optimization and Yield Prediction

Tamilarasi Kathirvel Murugan, Isshaan Singh, Divyansh Chawla, Raghav Jain

Published Jan 1, 2026• 1 citation

Publisher

DOI: 10.1007/978-3-032-17083-5_19

Related papers

AgriQML: A novel quantum machine learning framework for agriculture crop yield prediction

Deepa Bhadan, Thanjaivadivel M, R Lakshmana Kumar

Elsevier eBooksAug 14, 20260 citations

Harnessing Quantum Machine Learning Techniques for Enhancing Crop Productivity

S Sathya, Madhumathi Ramasamy

Lecture notes in networks and systemsJan 1, 20260 citations

A quantum-driven multi-stage framework integrating variational entanglement, reinforcement learning, and federated explainability for climate-resilient farming

Amreen Habibullah Khan, Dilip Kumar Jang Bahadur Saini, Tabassum H. Khan +3 more

The increasing constraints of climate change and data privacy necessitate high-efficiency, sustainable agriculture, which is causing a shift in the paradigm of Agro-informatics. Most classical agricultural models fail to capture genotype, soil chemistry, and climate dynamics connections. Latent interactions, essential to intelligent agricultural treatments and explainability, are lost in many data processing pipelines that use linear dimensionality reduction or black-box learning. This paper presents a quantum computing architecture for a revolutionary agricultural application, utilizing quantum encoding, topological learning, reinforcement optimization, federated intelligence, and explainability to highlight the importance of this vital field. In Quantum Variational Crop-Soil Entanglement Encoding, crop-soil interaction datasets are encoded into quantum state vectors using variational circuits, preserving high-order entanglement properties (fidelity > 0.96, entropy ~ 0.9). Quantum-guided agri-topological dynamics mapping transforms encoded states into permanent topological maps using a hybrid quantum-classical Topological Data Analysis to track climate-induced agri-system dynamics (r = 0.84 with the yield index). Field-level decisions using Quantum Reinforcement Learning for Precision Intervention policy mappings to relate topological states to interventions produce 16.2% normalized yield. Quantum Federated Learning for Distributed Farm Intelligence uses privacy-preserving, encrypted quantum policy gradients to enable learning across farms in varied locations, lowering communication by 42% and improving accuracy by 9.3%. Quantum Explainability through Entropic Intervention Attribution generates causal graphs of yield drivers with 89% confidence intervals using entropy-based attributions. This integrated framework enhances the knowledge preservation, policy accuracy, expandability, and trust of agricultural Artificial Intelligence systems, enabling quantum-accelerated, information-based, future-ready farming decision support systems.

Scientific ReportsNov 3, 20254 citations

Wireless Sensor-Driven Smart Agriculture Using Scalable Quantum Convolutional Models

Ladislau Mulyadi

Smart agriculture has emerged as a transformative approach to enhance crop productivity, resource efficiency, and environmental sustainability using advanced sensing and computational technologies. Wireless Sensor Networks (WSNs) enable real-time monitoring of agricultural parameters such as soil moisture, temperature, humidity, and crop health. However, traditional machine learning models struggle to efficiently process high-dimensional, noisy, and heterogeneous agricultural sensor data. This study proposes a Wireless Sensor-Driven Smart Agriculture framework using Scalable Quantum Convolutional Models (SQCM). The proposed model integrates wireless sensor data acquisition with quantum-inspired convolutional neural networks to enhance pattern recognition, predictive accuracy, and scalability. The quantum convolutional layers enable high-dimensional feature representation, while classical layers ensure efficient decision-making for agricultural optimization. The model is evaluated using agricultural sensor datasets, and performance is measured using accuracy, precision, recall, and RMSE for environmental prediction tasks. Experimental results demonstrate that the proposed quantum-enhanced architecture outperforms traditional machine learning and deep learning models in crop condition prediction and resource optimization. The framework is suitable for precision agriculture, automated irrigation systems, and smart farming applications.

International Journal on Advanced Computer Engineering and Communication TechnologyJun 4, 20260 citations

Harnessing quantum computing for smart agriculture: Empowering sustainable crop management and yield optimization

Chrysanthos Maraveas, Debanjan Konar, Dimosthenis K. Michopoulos +2 more

Computers and Electronics in AgricultureFeb 10, 202496 citations

Quantum Reinforcement Learning for Quantum Architecture Search

Samuel Yen-Chi Chen

This paper presents a quantum architecture search (QAS) framework using quantum reinforcement learning (QRL) to generate quantum gate sequences for multi-qubit GHZ states. The proposed framework employs the asynchronous advantage actor-critic (A3C) algorithm to optimize the QRL agent, which has access to Pauli-X, Y, Z expectation values and a predefined set of quantum operations. Our approach does not require any prior knowledge of quantum physics. The framework can be used with other QRL architectures or optimization methods to explore gate synthesis and compilation for various quantum states.

Aug 10, 202323 citations