An adaptation model for simulation of aging process

To analyze aging processes, a concept based on these hypotheses was utilized: maturation and aging are affected by external influences and are related to the system's ability to learn, to adapt to sudden environmental changes rapidly and survive. To simulate the maturation and aging processes of systems, individuals and populations, a perceptron to classify two notably different classification tasks was trained. After lengthy training to learn the first task, the magnitudes of the perceptron's weights increase. The presence of nonlinearity in the output of the perceptron causes a saturation of the cost function. Saturation reduces the capability to learn a new task rapidly. The aging curves obtained show a rise and fall character. Factors that can be utilized to control re-training curves were considered. A new model allows us to analyze the aging process as a natural phenomenon that helps populations to survive in everlastingly changing environments was also introduced.