A supervised neural network system is used for detection of tool breakage in milling operations. The cutting force signals presented are Restricted Coulomb Energy (RCE) type neural networks after encoding. The training period is very short compared to backpropogation neural networks. Also, the RCE networks indicate if their previous training to identify signals is satisfactory. The effectiveness of the encoding and convenience of the RCE networks are evaluated on simulated and experimental cutting force signals. The neural network is trained on simulated cutting force signals. The RCE network correctly categorized more than ninety-eight of the presented data sets after training, which include simulated and experimental cutting force data never before seen. The performance of the encoding method and the RCE networks are found to be acceptable for many tool breakage monitoring tasks of end milling operations.