Soil Nutrient Prediction Model in Hybrid Farming Using Rule-Based Regressor

Abstract

Agriculture is the supreme trade of India. Inefficiency and imprecise input controls such as nutrients of the soil, water, and usage of hazardous manure have caused devasting consequences to the biosphere. Various plant species are generated daily, however, they are deficient in all the necessary nutrients compared to the crop that is grown organically. To overcome this situation, an integrative hybrid approach is proposed in this paper, to combine both precision farming and organic farming which involves growing and fostering crops without the use of non-natural fertilizers and pesticides to elevate and enhance the quality and quantity of the crops. This paper proposes a machine learning (ML) model to predict nutritional values in Ballarat (Centella Asiatica) in both conventional farming and pro-biotic farming. This Hybrid system predicts the nutrient values such as nitrogen, phosphorus, and potassium with the supplement of banana peel powder prediction accuracy, mean absolute Error (MAE), Root mean squared error (RMSE), and R2 for potassium evaluate the performance of the model. The results reveal that random forest regressor performs well in probiotic farming with 91% and RMSE is 1.7475 and MAE is 0.6361 than decision tree regressor.