Predictive Analysis of Compost Combinations with Organic and Inorganic Materials for Crop Yield Optimization using Machine Learning

Sustainable agricultural practice faces major obstacles in optimizing compost quality through agricultural residue use while reducing chemical inputs. Traditional composting involves adding high levels of urea to achieve proper carbon-to-nitrogen (C:N) balance and speed up the process of maturation. An excessive amount of urea application results in nitrogen escaping as vapors, drives up costs, and causes environmental deterioration. The research develops a machine learning predictive model that examines compost combinations of agricultural wastes along with biofertilizers at reduced urea proportions. Tamil Nadu Agricultural University data combined with government agrometeorological records led to the creation of more than 1,48,000 compost combinations. We studied how inorganic compost made from urea variation and organic compost based on biofertilizer addition performed at predicting rice harvest results through analyses of NPK components and C:N ratios. The best compost options were selected through Euclidean distance matching analysis of actual rice yield data before ensemble learning models finalized their characteristics by ranking features. Environmental sustainability improved when microbial biofertilizers Azotobacter and PSB were added to composts since the resulting yields matched those of urea-based methods. The presented research calls for decreased chemical use alongside for environmentally friendly data-based farming systems that manage nutrients.