Asymmetric rolling is a process in which the radius of the rollers, the peripheral speed of the rollers or roller friction conditions are not as the same as the sheet. Both the rolling force and the sheet exit curvature in the process depend on the values of eight parameters including inlet thickness of sheet, thickness reduction, radius of the upper and lower roll, work rolls speed ratio, friction factor of the upper roll, friction factor of the lower roll and mean yield shear stress of the sheet. In this thesis, the value of the rolling force which must be minimized is considered as the cost function of the problem, the curvature index of the sheet which must be close to one as the nonlinear and main constraint of the problem and the range of the design parameters as the linear constraint. In order to establish a relationship and the mathematical model of the optimization process, a function must be extracted in which the rolling force is stated relating to the forementioned eight parameters and also a function in which the Curvature index of the sheet is stated in relation to the forementioned eight parameters. For this purpose, the data and results which have been obtained from the numerical solution of analytical equations of asymmetrical rolling process in different conditions, which show the values and trend of the change of the rolling force and also the index of curvature, are used separately as the input data for training two artificial neural networks. One of the networks is used for the extraction of cost function and the other for the extraction of curvature constraint function. Then, in order to implement the optimization process, a program with the quasi newton algorithm is codified and after its implementation, the optimized values of effective parameters in asymmetrical rolling are obtained. With these optimized solutions, the sheet exit curvature is very low and on the other hand the rolling force is minimized. Then, for a better understanding of the process of optimal design and the effective value of the design parameters, steps of the iterative process such as sensitivity analysis of the objective function and constraints, changing in the bounds of constraints and drawing the iterative diagrams are implemented. At the end, the optimized results are compared with the results of other researches and with the aid of the sensitivity analysis, determination of the influence value of each effective parameter on the rolling force value, and exit curvature of the sheet, the optimized results of this research and the reasons of their being more optimum are analyzed. Keywords: Interactive Optimization, Asymmetric Rolling, Rolling Force, Sheet Exit Curvature