IMPLEMENTATION OF AN MS EXCEL TOOL FOR BACKPROPAGATION NEURAL NETWORK ALGORITHM IN ENVIRONMENTAL ENGINEERING EDUCATION

Abstract

This paper presents the implementation of an MS Excel tool for backpropagation neural networks in environmental engineering education. A number of test cases, Darcy friction factor and oxygen solubility, were also provided to test the performance of the tool. Relative mean square errors and coefficients of determination for Darcy friction factor were calculated as 1.53*10(-3)+/- 0.59*10(-3) and 0.9983 +/- 0.0003 while they were calculated as 4.78*10(-4)+/- 2.33*10(-4) and 0.9998 +/- 0.0000 for oxygen solubility. Results suggested that the tool's performance is satisfactory. The tool produces satisfactorily fast and reliable results and can be used for environmental education in undergraduate/graduate level.