Machine Learning: Crop Yield Prediction and Fertilizer Recom

This project predicts the estimated crop yield (Lint Yield of Cotton) based on input parameters (US State, Area Planted, Area Harvested, N-P-K Acreage % and N-P-K Quantity in Pounds/Acre). The data set analyzes historical data of 14 states in the US since 1964 and also recommends the fertilizer quantity to be used. The prediction model is based on Gradient Boosting Regression (90.1% R2 value) with K Folds Cross Validation (5). Fertilizer Recommendation uses the Multi-Variate Regression algorithm.