Genetic algorithms and back-propagation: a comparative study

Both genetic algorithms and backpropagation are search and optimization techniques. They are used to optimize complex and nonlinear systems. However, no comparative study has been carried out or found in the literature so far, in order to investigate the relative performance of both algorithms. In this paper, a multi-resolutional dynamic genetic algorithm (MRD-GA) is described and used to model a gas furnace. Also, a back-propagation algorithm is developed and used to model the same system. From the modeling results, a comparison is carried out to show the strengths and limitations of each paradigm