Using Gravitational Search Algorithm for Solving Nonlinear Regression Analysis

Evolutionary algorithms (EAs) provide a framework for dealing with a variety of large-scale multi-objective problems (MOPs) in the field of evolutionary algorithms, when applied to different problem types. Although computational strategies for dealing with nonlinear regression problems are difficult to apply, we used the gravity search algorithm and combined it with these regressions to estimate and interpret the parameters. Estimation parameters for nonlinear regression and gravitational search algorithm (EPNGSA), enabling us to access them consistently. Estimating the nonlinear estimation parameter using general estimating equations. It is necessary to use Chebyshev's strategy in the leader recruitment procedure, which leads to tackling (MOP) based on the Gravity Search Algorithm (GSA) and at the same time may lead to quick results. When building a leader library, the concept of dominance is crucial because it allows leaders, they choose to include less dense regions, thus producing an estimated Pareto front with a large diversity, which reduces global optimization challenges. The used and new method showed its effectiveness and was closer to the solution compared to other algorithms. This result was obtained using six standard nonlinear functions. GSA appears to be more productive than both Practical Swarm Optimization PSO and Bat Algorithm BAT.