Water Quality Prediction of the Yamuna River in India Using Hybrid Neuro-Fuzzy Models

The potential of four different neuro-fuzzy embedded meta-heuristic algorithms, particle swarm optimization, genetic algorithm, harmony search, and teaching–learning-based optimization was investigated in this study estimating the water quality Yamuna River Delhi, India. A cross-validation approach employed by splitting data into three equal parts, where models were evaluated using each part. main aim to find an accurate prediction model for River. It is worth noting that hybrid LSSVM methods have not been previously compared issue. Monthly parameters, total kjeldahl nitrogen, free ammonia, coliform, temperature, hydrogen, fecal coliform considered as inputs chemical oxygen demand (COD). performance predicting COD with classical least square support vector machine (LSSVM) methods. results showed higher accuracy when temperature used inputs. Hybrid improved root mean error 12% 4%, respectively. optimized search provided best lowest (13.659) absolute (11.272), while highest computational speed (21 24 min) other models.