Viability of two adaptive fuzzy systems based on fuzzy c means and subtractive clustering methods for modeling Cadmium in groundwater resources

The Adaptive Neuro-Fuzzy Inference System (ANFIS) combines the strengths of both Artificial Neural Networks (ANNs) and Fuzzy Logic (FL) into a single framework. By doing so, it allows for quicker learning adaptable interpretation capabilities, which are useful modeling complex patterns identifying nonlinear relationships. One significant challenge in assessing water quality is difficulty time-consuming nature determining various factors that impact it. Given this situation, predicting heavy metal levels groundwater resources, urban rural, essential. This paper investigates two methods, ANFIS-FCM ANFIS-SUB, to determine their effectiveness Cadmium (Cd) resources. ‏The parameters be considered are: dissolved solids (TDS), electroconductivity (EC), turbidity (TU), pH were assumed independent variables. A total 51 sampling location used with resource develop fuzzy models. For evaluating performance ANFIS-SUB models, three different criteria including correlation coefficient, root mean square error, sum error comparing model outputs actual outputs‏.‏ Based on obtained results from scatter plots predicted value by ANFIS- FCM determination coefficient (R2) data, test train sets equal 0.978, 0.982, 0.993 0.983, 0.999 0.998 respectively. result proved Cd predictions implemented was significantly close measured all experimental data R2 0.983. compared found provided slightly higher accuracy than model. Also, comparison between indicated have strong potential estimating metals high degree accuracy.