In the present study, the EC and TDS quality parameters of Karun River were modeled using data mining algorithms including LSSVM, ANFIS and ANN in three Molasani, Ahvaz and Farsiat hydrometric stations. Nine different combinations of inputs including, Cl-1, Ca + 2, Na+ 1, Mg+ 2, K+ 1, CO3, HCO3, and SO4 with discharge Q were used as the algorithm inputs. Modeling results showed that Na+1, Cl-1 and Cl+2 have the most influence on the modeling of EC and TDS parameters. The LSSVM algorithm is most accurate in estimating both EC and TDS parameters. On the other hand, among the stations studied, the most precision for EC modeling is related to Ahvaz station, which has a 16% higher coefficient of determination. For TDS, the highest accuracy was related to the Molasani station with 36% higher coefficient of determination than the other two stations. In the next step, the EC and TDS parameters are modeled based on the four parameters includes Na+1, Cl-1, Ca+2, and Q, and with lags timelag times equal to zeroszero to three months. The results showed that EC parameter in Ahvaz station with zeros monthzero months' lag time had the best accuracy. So that, the MAE, RRMSE and R2 for the mentioned station are equal to 72.82, 0.15 and 0.98, respectively. It is worth noting that comparing the EC and TDS time series from LSSVM and observed time series, indicates totoo high accuracy of LSSVM in estimating of oscillationoscillation and peak EC and TDS parameters inat varying timetimes. Therefore, in the present study, the methods and models can be a useful tools for decision-making in the quality management of Karun basin.