Hardy Weinberg Principle Automation : A Comparative study of Neural network and Fuzzy based accessibility model for the study of Evolution

Major diversity in population, heredity, and varied survival rates contribute to evolution in organism populations. A common technique to study evolution in population is to investigate the variation of the population's allele frequencies (genetic frequency) from one generation to the other. In plain terms this means that instead of focusing on two main parental species evolution should be studied by investigating the entire population. The most effective method is to analyze the variation in allele frequency in the first generation of the population and then predict the corresponding variation in the future filial generations. Hardy-Weinberg Equilibrium model equations in Ecology and Sustainability are used in general to predict a population's allele frequency in the future. As the population model becomes complicated these ecological equations become too difficult to realize and likewise the prediction of the population of the species for the future generation. For example, considering the re-fragmentation of forest landscape. This paper presents two unique methods that utilizes computational intelligence (CI) techniques to automate the Hardy-Weinberg principle by integrating Engineering and Ecology. The automation is done using two CI techniques, which are the Mamdani Fuzzy Inference System and Back Propagation Theory. The Mamdani FIS is tuned using MATLAB fuzzy inference toolbox. The back propagation training model is deployed using python which utilizes Keras and Tensorflow for predicting evolution. To prove the efficiency of the trained models a standard allele frequency database named “ALFRED” is used.