Micronutrient Classification in IoT Based Agriculture Using Machine Learning (ML) Algorithm

Globally agriculture seems to be the economic field
demographically w -economic structure. At present peoples are not aware of crop uncertainty of food, due to seasonal climatic environments based on the cultivation techniques are being changed against the basic assets such as air, water and soil. For this reason, Machine Learning (ML)
techniques are the best choice for agriculture and evaluated to predict agriculture growth. This paper proposed horticulture IoT monitoring sensor board to develop an IoT architecture in agriculture industry for monitoring the Micro and Macro Nutrients of S oil and analyze various soil parameters present in the Thiruvarur District in Tamil Nadu. The framework assists in making correct decisions based on data gathered from the IoT sensors which are stored on the server and analyzed using ML algorithms. The ML model is used to classify the dataset based on the threshold value of micro and macro nutrient obtained from National Food S ecurity Mission (NFS M). The evaluation can be done by using various ML classification algorithm such as Naive Bayes (NB), Logistic Regression (LR), Random Tree (RT) and K-
Nearest Neighborhood (KNN). The classification method is
compared and evaluated through accuracy, Root Mean S quare
Error (RMS E), Relative Absolute Error (RAE), Mean Absolute
Error (MAE), Root Relative S quared Error (RRS E). The KNN
classifier attains lower MAE, RMS E and RRS E value of 0.2398, 0.3908, and 94.1845 and outperforms the other three classifier but RT algorithm attains lower RAE value of 66.24 than KNN.