The current brain stimulators are programmed by trial and error. This is extremely time consuming (takes hours for each session and needs to be repeated over months) and subjective (depends on the testing neurologist). There is a need for systems that can optimize the programming parameters based on the response of the brain in an automated and objective manner. The system/method proposed here probes the brain to get the default neural response for an individual patient and infer the optimal stimulation frequency automatically in a closed-loop fashion.