Showing 2 results for Regression Models.
M Motamednia , S.h.r Sadeghi, H Moradi, H Asadi ,
Volume 14, Issue 52 (7-2010)
Abstract
An extensive data collection on precipitation and runoff is required for development and implementation of soil and water projects. The unit hydrograph (UH) is an appropriate base for deriving flood hydrographs and therefore provides comprehensive information for planners and managers. However, UH derivation is not easy job for whole watersheds. The development of UH by using easily accessible rainfall data is then necessary. Besides that, the validity evaluation of different statistical modeling methods in hydrology and UH development has been rarely taken into account. Towards the attempt, the present study was planned to compare the efficiency of different modeling procedures in hydrograph and 2-h representative UH relationship in Kasilian watershed with concentration time of some 10h. The study took place by using 23 storm events occurred during four seasons within 33 years and applying two and multivariable regression models and 36 variables. According to the results, the median of estimated errors in estimation of 2-h UH dependent variables for verification stage varied from 37 to 88%. The results verified the better performance of cubic and linear bivariate models and logarithm-transformed data in multivariable model as well. The efficiency of multivariable models decreased when they were subjected to principle component analysis. The performance of backward method was frequently proved for estimation of dependent variables based on evaluation criteria, whereas the forward was found to be more efficient for time-dependent factors estimation.
R. Rezae Arshad, Gh. Sayyad, *, M. Mazloom, M. Shorafa, A. Jafarnejady,
Volume 16, Issue 60 (7-2012)
Abstract
Direct measurement of soil hydraulic characteristics is costly and time-consuming. Also, the method is partly unreliable due to soil heterogeneity and laboratory errors. Instead, soil hydraulic characteristics can be predicted using readily available data such as soil texture and bulk density using pedotransfer functions (PTFs). Artificial neural networks (ANNs) and statistical regression are two methods which are used to develop PTFs. In this study, the multi-layer perceptron (MLP) neural network and backward and stepwise regression models were used to estimate saturated hydraulic conductivity using some soil characteristics including the percentage of particle size distribution, porosity, and bulk density. Data of 125 soil profiles were collected from the reports of basic soil science and land reclamation studies conducted by Khuzestan Water and Power Organization. The results showed that MLP neural network having Bayesian training algorithm with the greater coefficient of determination (R2=0.65) and the lower error (RMSE =0.04) had better performance than multiple linear regression model in predicting saturated hydraulic conductivity.
