Ensemble Learning Techniques for Rice Nutrient Disease Deficiency Detection and Prediction Analysis

Abstract

Rice is a vital food source and its nutritional composition, including essential minerals and vitamins, significantly impacts human health. Understanding nutrient deficiencies and diseases in rice is crucial for promoting healthy and sustainable agriculture and preventing related health problems. Rice grain mostly suffers from production issues triggered by nutrient imbalances like potassium, phosphorus, and nitrogen. Generally, nutrient deficiencies in rice plants show stimulation due to differences in leaf colour. Leaf features provide nutrient shortage classification of colour and shape. This study presents ensemble learning to classify rice crop nutrient deficiencies. The datasets were taken from the Kaggle data source. It consists of hundreds of rice leaf images, it can be divided into different classes. They can represent deficiencies in potassium, nitrogen, and phosphorus. This paper concentrates on applying ensemble learning to predict and analyse outcomes. This paper focused on applying machine learning techniques to analyse and predict outcomes using different models, including Linear Regression for continuous predictions. Random Forest for robust classification. XGBoost for high-accuracy predictions. K-Nearest Neighbours (KNN) for pattern recognition. By testing multiple models and comparing their performance, we identified the most successful algorithm for our dataset.