A numerical controller performs machine learning to calculate tailstock pressing force to stabilize behavior of a workpiece during lathe machining and acquires, as state information, current tailstock pressing force. Then, the numerical controller acquires, as determination data, current vibration or displacement of the workpiece, or power consumption by tailstock pressing force, updates a value function based on the acquired determination data, and outputs an adjustment value of the state information based on the updated value function.

A numerical controller performs machine learning to calculate tailstock pressing force to stabilize behavior of a workpiece during lathe machining and acquires, as state information, current tailstock pressing force. Then, the numerical controller acquires, as determination data, current vibration or displacement of the workpiece, or power consumption by tailstock pressing force, updates a value function based on the acquired determination data, and outputs an adjustment value of the state information based on the updated value function.

A method for setting or monitoring operating parameters of a workpiece processing machine having a tool holder and means for moving a workpiece and the holder relative to one another along a first axis. During milling of the holder fitted with a workpiece and during application of the tool to a workpiece, values for at least one measured variable are registered and recorded. The variables include an axial force acting in a direction parallel to the first axis, a torque relative to the first axis or to an axis oriented parallel to the first axis, and bending torques or bending torque components according to direction and amount. The values are used to set the operating parameters with respect to an extended service life of the tool in conjunction with a processing time falling below a maximum machining time, monitoring tool wear or a machine error.