Estimation of models for cumulative infiltration of soil using machine learning methods

Authors:
Publisher website
Google Scholar
Abstract
Knowledge of cumulative infiltration of soil is necessary for irrigation, surface flow, groundwater recharge and many other hydrological processes. In the present study, the Support Vector Machine (SVM), artificial neural network (ANN) and adaptive Neuro-fuzzy inference system (ANFIS) were employed to estimate the cumulative infiltration of soil. For this study, a data set containing 106 experimental observations were analyzed. Out of 106, 70 % of data was selected for preparing different algorithms whereas rest 30% data was selected to test the models. The models accuracy was depended upon the two performance evaluation parameter which is correlation coefficient (CC) and root mean square error (RMSE). Results of performance evaluation parameters suggest that triangular membership function-based ANFIS model works well than SVM and ANN models. While SVM and ANN models also give a good estimation performance. Sensitivity analysis concludes that the parameter, time (t) is the most influencing parameter for the modeling of cumulative infiltration of soil for this data set.
Funding Information
  • no (nil)

This publication has 42 references indexed in Scilit: