In this thesis an attempt has been made to minimize the flatness defects in rolled strips. In cold steel rolling, strip shape is crucial to product quality. For modern rolling mills, there are a number of different ways to control the strip shape, including adjusting the side depression, bending the work rolls and axial shifting the middle roll. However, these controls are not independent and hence, must be used with great care. In recent times, models based on data mining techniques are being increasingly used in the areas of mechanical property prediction and shape defect analysis in the steel industry. Very few well-established models based on either mathematical or data mining techniques, could be developed into hybrid models for hot rolling. Compared to these conventional digital computing techniques, models based on a combination of artificial neural network and genetic algorithm are advantageous because of their special features, such as massive parallel processing, distributed sorting of information, robust operation after training, generalization, ability to handle both categorical and continuous variable and adaptability to new information. At first step, a premier model based on Slit Beam Method (SBM) is developed to consider elastic deformation of work roll and stand. After verifying new model, an artificial neural network model is prepared with the results of the latest SBM model. Finally finding the most optimum process parameters is performed using a novel hybrid model with the combination of a predictive artificial neural network (ANN) and a genetic algorithm (GA). This work introduces a Neural Network application to a plate mill to improve the model’s prediction ability for rolling force. Since thickness accuracy is highly related to the rolling-force precision, its improvement is very important. Conventional methods with simple mathematical models and a coarse learning scheme are not sufficient to maintain a good prediction ability for the rolling force because the rolling force variable has very nonlinear and time-varying characteristics. These problems are alleviated when an on-line adaptable Neural Network is applied instead. Basically, the Neural Network is capable of compensating the nonlinear model Deficiency. GA is meant for the optimization of few selected process parameters affecting the flatness of the strip by a proper fitness function and forms a closed loop with the ANN. The implementation results are compared with simulation results to verify viability of proposed method. Different practical conditions and software optimization techniques are discussed in details. At the end, there is a summary of thesis and suggestions for further studies. Keywords: Flatness, Profile, Rolling control, Artificial neural network (ANN), Genetic algorithm (GA), Optimizatio