Finding of Maturity Level of Crops for Harvesting using IOT and ML Based Structured Framework

IoT designs make it possible to collect data for vast and remote agricultural regions, which may then be used to use machine learning algorithms to anticipate crops. In order to establish the suggested crop, recommendations are based on factors including temperature, humidity, rainfall, pH, N, P, and K. The dataset includes 2200 occurrences and 8 characteristics. For different combinations of these attributes, about 22 alternative crops are suggested. Using specific machine learning algorithms in WEKA, supervised learning techniques are used to create an ideal model. The rules-based classifier JRip, the decision table classifier, and the multilayer perceptron are the machine learning methods selected for classification. Establishing a model that can forecast high-yield crops and facilitate precision agriculture is the main goal of this case study. IoT technologies and crucial agricultural metrics are included in the suggested system modeling. A weighted average receiver operator characteristic of one and an accuracy of 98.2273% are shown by the performance evaluation of the chosen classifiers. 8.05 seconds is the maximum amount of time needed to construct the model.