One such crop that is widely farmed around the world is turmeric. Crop output is significantly impacted by the plant disease. The most typical plant diseases are rhizome rot, leaf spot, and leaf blotch. Rhizome rot is one of the diseases that has been documented to be particularly harmful to the production of turmeric, with a 60% crop loss rate. In order to address the multi-class unbalanced rhizome rot dataset classification, adaptive boosting (AdaBoost) in conjunction with a potent ML classifier (Random Forest) is suggested in this study. AdaBoost builds a strong classifier by integrating many sub-classifiers based on weights. Using Random Forests as the first stage classifier and AdaBoost as the second stage classifier to determine which class the illness sample falls into is a novel, robust, and more accurate method offered. To assess the robustness of the suggested hybrid AdaBoost approach, three different datasets are used. The suggested technique enhances both accuracy and stability for the unbalanced rhizome, according to experimental findings, which are compared to various state-of-the-art algorithms (kNN, SVM, RF, kNN-AdaBoost, and SVM-AdaBoost).RFAdaBoost has a 92% f1_score and yields the lowest root mean squared value. When the hybrid algorithm's stability is assessed using the variance, g-means metric value, and Matthew’s correlation coefficient (MCC), RFAdaBoost performs better than the other cutting-edge methods.