Crop Yield Prediction of Cotton Using Optimization Technique

Farmers still use ancient methods in nations like India. Without knowledge of the soil's actual productivity or likelihood of supporting the crop being planted, crops are planted based on knowledge learned from prior experiences. As a result, the net profit and harvest do not reach their full potential. A soil sample is taken and sent to a lab for a variety of physical and chemical analyses as part of the manual approach currently in use for soil testing. As humans are involved, there is a higher potential of mistakes occurring, and farmers may receive inaccurate reports. The necessity to automate the process results from this. The purpose of soil testing is to identify the soil's fertility factor and, consequently, the best crop to farm for maximum yield. This prompted the development of software that estimates the soil's amounts of nitrogen (N), phosphorus (P), and potassium (K) after first calculating several soil fertility variables like moisture, electrical conductivity, pH, and others. With the aid of machine learning and gradient descent optimization technique for classification algorithms, the dataset is examined, and a forecast of the best crop to plant is predicted.