Metaheuristic approaches are inspired mainly based on the order and rules of natural organisms. Today, these approaches have been widely used in various branches. According to the importance of forecasting, understanding the methods of earnings management forecast can provide useful information for the stakeholders. The variety of factors obtained due to the results of linear patterns used for measuring earnings management has caused investors to hesitate the reported earnings quality. Therefore, the purpose of this research is to provide an optimal templete for earnings management forecast. In the first step, the origin linear model is optimized using the pattern of neural networks then Particle Swarm Optimization and Imperialist Competitive Algorithms are used to optimize the pattern more. The sample consists of 620 firms listed in Tehran Stock Exchange during the years 2010 to 2015. The results indicate usefulness and positive impact of panel data methods on the performance of earnings management forecast. The findings also show a significant difference between usefulness of the linear and nonlinear methods. In other words, using algorithms in earnings management forecast, the prediction accuracy increases with the elimination of inefficient variables. In addition, findings indicate a better and more suitable performance of Imperialist Competitive Algorithm than other patterns in the efficiency of the management variables with accuracy (95/8%).