Crop recommendation using machine learning

Abstract

Agriculture serves as the cornerstone of India’s economic expansion, constituting the primary income source for a significant proportion of its populace, encompassing both those directly engaged in agricultural activities and those who depend on it indirectly for their livelihoods. Therefore, it is essential for farmers to make the correct option possible when cultivating any crop so that the farmer can make maximum profit from the agriculture field. To make the agriculture sector profitable, one of the technologies that may be used in this age of rapid technological improvement is known as machine learning (ML). In this research paper, various ML algorithms, such as logistic regression (LR), decision trees (DT), Light GBM, and random forest (RF), have been utilized to analyze a dataset. The primary objective is to predict the most suitable crop based on soil attributes such as (nitrogen, phosphorous, potassium) NPK content, humidity, temperature, soil pH level, and rainfall. Out of this random forest and Light GBM comes with great accuracy whereas decision tree and logistic regression have less accuracy. In addition, ML algorithms will likely find applications in a variety of agricultural subfields in the near future, including the diagnosis of plant diseases, the selection of soil types, and the forecasting of retail pricing.