ABSTRACT Grape disease is one of the most common diseases that impacts grapevines, affecting both the yields of the plants and the quality of the fruits that are harvested. Currently, fungicide treatments are often used throughout the season to combat the disease. In terms of public health and the environment, fewer treatments are necessary. This issue could be solved by identifying vineyards that are more likely to suffer severe attacks in the spring and treating them only with fungicidal treatments. Computers, categorization, bioinformatics, marketing, healthcare, gaming and industry are just a few of the many areas that have made use of machine learning in the past several years. These technologies are used to identify high‐quality grapes for export after they have been photographed and pre‐processed. Accurate illness detection and disease preventive management methods are critical for increasing quality and productivity. This proposed study aims to effectively predict the quality of grape yield and perceive illnesses such as powdery mildew and downy mildew. Initially, sensors located on farms are used to gather climate data. Then, the input data is pre‐processed using min‐max normalization and a one‐hot encoding method to remove the unwanted data. Four distinct machine learning classifiers are then employed on the pre‐processed input: K Nearest Neighbours (KNN), Logistic Regression (LR), Multinomial Naive Bayes (MNB) and Bernoulli Naive Bayes (BNB). A variety of performance measures are used to assess the performance of the proposed model. In terms of disease prediction, the KNN classifier outperforms with 82% accuracy on plots 2 and 3. Additionally, plots 2 and 3 have a yield prediction accuracy of 98%. The results obtained are more efficient than other existing models.
Sinha et al. (Sun,) studied this question.