COA Based Adaptive Neuro-Fuzzy Inference System: Machine Learning Approach for Crop Yield Prediction to Promise Agricultural Sustainability

Accurate early estimation of crop yield is vital for effective agricultural planning, trade policy formulation, and enhancement of farmers’ income. This study presents an optimized Adaptive Neuro-Fuzzy Inference System (ANFIS) integrated with the Chimp Optimization Algorithm (COA) to achieve reliable and precise crop yield prediction. The model was trained on multi-year datasets encompassing crops such as wheat, urad dal, turmeric, sugarcane, rice, ragi, green gram, maize, linseed, jowar, and groundnut. Experimental outcomes indicate that the proposed COA-ANFIS approach achieves outstanding performance with 97.65% accuracy, 97.85% sensitivity, 97.74% specificity, 97.85% precision, and an F-measure of 97.54%, surpassing conventional ANFIS, SVM, and ANN models. Additionally, it demonstrates reduced execution time and minimal memory usage. Forecast analysis for 2022–2031 predicts maximum sugarcane yield in 2027 and minimum sunflower yield in 2029, confirming the model’s capability as an efficient and intelligent tool for predictive agricultural analytics.