Potato Leaf Disease Detection: Classical ML with Future Quantum and RL Directions

Roshini N

Abstract: Plant disease detection is essential for improving crop productivity and agricultural sustainability. Machine learning enables automated classification of leaf diseases, reducing reliance on manual visual inspection. This study experimentally evaluates K-Nearest Neighbors (KNN) and Random Forest (RF) for classifying potato leaf diseases using a balanced subset of the PlantVillage dataset. Images were preprocessed through resizing, normalization, and flattening prior to classification. Random Forest outperformed KNN across all metrics, achieving 88.00% accuracy versus 75.33% for KNN. Beyond the experimental study, this paper reviews recent advances in Quantum Machine Learning (QML) for agricultural disease detection and proposes a Reinforcement Learning (RL)-based framework as a conceptual future extension for adaptive disease diagnosis. The RL framework is not experimentally implemented in this work. The results establish reliable classical baselines while highlighting future directions for QML and RL in precision agriculture.

Keywords: Plant Disease Recognition, Precision Agriculture, Random Forest, K-Nearest Neighbors, Quantum Machine Learning, Reinforcement Learning, PlantVillage Dataset

How to Cite?: Roshini N, "Potato Leaf Disease Detection: Classical ML with Future Quantum and RL Directions", Volume 15 Issue 8, August 2026, International Journal of Science and Research (IJSR), Pages: 767-769, https://www.ijsr.net/getabstract.php?paperid=SR26811165236, DOI: https://dx.doi.org/10.21275/SR26811165236

Download Citation: APA | MLA | BibTeX | EndNote | RefMan