Harmony search (HS) has been applied in various fields and presented good results. However, since HS has several drawbacks (i.e., parameter settings, solution stagnation), various improved versions of HS developed to overcome these drawbacks. In this study, a new improved HS is proposed called inter-ethnic harmony search (IeHS). IeHS is developed to mimic the concept of the reconciliation of various ethnicities from Turkish history, and this algorithm considers the self-adaptive parameters and decision variables setting, balancing local and global search, overcoming the solution stagnation problem using the historical concept (i.e., Millet and Jannissary). To verify the performance of IeHS, the mathematical benchmark functions from CEC 2014 are applied and compared representative improved version of HS using performance indices.