Material removal rate in numerical controlled machining, especially in thin rib machining, is limited by the onset of tool path error and chatter. In this paper we propose that it is possible to increase the metal removal rate while maintaining the required dimensional accuracy and surface finish by modeling the cutting process as stochastic. With this improved modeling it will be possible to predict the characteristics of finished surface more accurately. A methodology to identify the system parameters and the design of an optimal controller to minimize stochastic tool path error in thin rib machining is presented.