Food intake, bodyweight and appetite are controlled by a "web of hormones." Recently, from this web of hormones, several hormones have been revealed and investigated with different degrees of success: a key one is ghrelin. Accordingly, ghrelin is an orexigenic (i.e., appetite stimulant) hormone; in fact, the only one of its kind, a peripheral hormone that can influence, centrally, one's propensity to start a meal. On this work, we shall present a problem in parameter estimation using evolutionary algorithms in conjunction with local search, what we have called herein hybrid algorithms (global search + local search); additionally, we apply artificial neural networks (feedforward neural network) for supporting the numerical simulations (what we have named "fake data"). Moreover, we present a mathematical model for ghrelin partially published elsewhere by the same authors; in addition, we have confronted the model mathematically with in vivo data via parameter estimation and got promising results for the novel mathematical formulation for ghrelin dynamics. Thus, our aim is showing that our algorithms can be imperative for fitting the current and future versions of the model. Notwithstanding the parameter estimation was unable to model precisely the experimental data, most likely due to physiological details still unclear in the medical literature, it generated an optimized curve relatively close to the experimental data, leaving promising results for future investigations.