The systems and methods can automatically determine a patient-specific set of stimulation parameters using optimization for exploring and/or programming a neurostimulation device, e.g., deep brain stimulation (DBS). In one implementation, the method may include receiving patient data and set of stimulation parameters associated stimulation delivered to a patient by a neurostimulation device. The method may include determining one or more objective metrics using the patient data for each set of stimulation parameters. The one or more objective metrics may be a quantitative value that represents a rating or score of tremor severity and/or side effect severity. The method may further include generating a patient specific response model using the one or more objective metrics and the associated set of stimulation parameters. The method may also include applying an optimization algorithm to the generated response model to determine a candidate set of one or more stimulation parameters.