the Importance of Teaching Quality Factors from Student's Viewpoints by Applying Artificial Neural Network Sensitivity Analysis

Students are considered as the main informatic source of quality and effectiveness of education. By determining the importance of teaching quality factors from student's viewpoints, The population of this study 3097 student's including student's of Agriculture and Natural Resources at the University of Tehran (2027 M.Sc. and Ph.D students) and Faculty of Agriculture at Ferdowsi University (1070 M.Sc. and Ph.D students). this research has studied the viewpoint of graduate and post graduate student's of Agriculture and Natural Resources at the University of Tehran (269 M.Sc. and Ph.D students) and Faculty of Agriculture at Ferdowsi University (202 M.Sc. and Ph.D students) by applying artificial neural networks. Several ANNs were designed, trained and generalized for predicting the teaching quality from student's viewpoints in UTCAN, among these networks, a network with 7- 4- 4- 1 architecture had the highest determination coefficient (R2 = 0.855) and the lowest values of Mean Square Error (0.36) and Mean Absolute Percentage Error (13.6%). A network with 7- 38- 36- 34- 1 architecture has the best results for (R2= 0.915), Mean Square Error (0.07) and Mean Absolute Percentage Error 2.87%) between the actual and modeled values of teaching quality from student's viewpoints in FAFU. The results of sensitivity analysis of ANN modeling for UTCAN showed that communication skills and dominance upon content had the highest sensitivity at UTCAN teaching quality from student's viewpoints, respectively and dominance upon content and communication skills had the highest sensitivity at FAFU teaching quality from student's viewpoints.