Abstract:

Abstract The designs of the Internet of Things make it feasible for us to collect data for large agricultural areas that are located in remote parts of the world. Our machine learning system can then utilize this data to produce crop predictions. The crop that should be recommended is based on the following factors: nitrogen, phosphorus, potassium, temperature, humidity, and rainfall. On the basis of these considerations, recommendations are made. The data collection has a total of 2200 occurrences as well as 8 attributes to go along with them. There are around 22 different plant species that might be offered for each of the 8 potential combinations of traits. By making use of the supervised learning strategy and using some of the machine learning algorithms that are accessible in WEKA, it is possible to create the best possible model. For the purpose of the classification procedure, the multilayer perceptron rules-based classifier JRip and the decision table classifier were selected as the candidates for the machine learning algorithms that would be implemented. The major objective of this case study is to, by the time it's finished, develop a model that not only predicts the crop to have a high yield but also provides guidance for precision agriculture. Both the Internet of Things and the essential metrics that are needed in agriculture are taken into consideration in the model that has been provided for the system. It has been established that the performance that was assessed by the classifiers that were selected has a value of 98.2273%, that the weighted average Receiver Operator Characteristics value is 1, and that the maximum amount of time that is required to generate the model is 8.05 seconds. The agricultural sector contributes significantly to the overall economy. It is very necessary for the upkeep of a balanced ecology. People are dependent on diverse agricultural products to some degree in almost every facet of their lives. This is especially true in terms of food and water. Farmers need to discover strategies to adapt to shifting weather patterns while also satisfying the increased demand for more food of higher quality. If the farmer wants to boost the output of crops and ensure their healthy development, he or she has to be aware of the weather conditions. Because of this information, the farmer will be able to make an informed decision regarding the kind of crop to cultivate, taking into account the myriad of environmental factors. Monitoring the crop in real-time allows Internet of Things (IoT)-based smart farming to make the whole agricultural system more efficient. It monitors a variety of elements such as humidity, temperature, and soil, among others, and provides a real-time observation that is utterly transparent. The purpose of using machine learning in the agricultural sector is to enhance both the output and quality of the crops produced in this industry. When applied to the sensed data, the use of appropriate algorithms may assist in the selection of suitable crops.

Keywords: