ABSTRACT
The 'best so far' individual or the best set of parameters that correspond to the minimum/maximum objective function is recorded over the generations. Genetic algorithms (GAs) operate on a coding of the parameters, rather than on the parameters themselves. Methods of finding the optimal solution include 'hill climbing', 'simulated annealing', and 'genetic algorithms'. GAs consist of the following steps: The questions that need clarification include how to create chromosomes, what type of encoding to use, and how to select parents for crossover. At this stage, some prior information the feasible range of values of the parameters could help accelerate the search process. As in any standard evolutionary algorithm, in order to optimize its fitness value during the evolving process, the trees in genetic programming (GP) are dynamically modified by genetic operators. GP has the unique feature that it does not assume any functional form of the solution, and that it can optimize both the structure of the model and its parameters.
