Pathological tremor is an involuntary, rhythmic movement that can inhibit the ability of a person to perform everyday tasks. Recent research explores mechanical means of tremor suppression as an alternative to drugs and surgery. However, traditional control methods also suppress voluntary movements due to the close proximity of tremor frequency and the frequency range of typical voluntary motions. Therefore, the controller must identify and suppress the tremor torque with minimal influence on voluntary movement. In addition to the control design, the actuator plays a critical role in the performance and potential for clinical implementation of a tremor suppression system. Dielectric elastomers offer unique actuation capabilities due to their low stiffness compared to traditional engineering actuators. In particular, dielectric elastomers have similar mechanical properties as human tissue, making them ideal for actuation of the human body. This work applies an adaptive notch filter algorithm for vibration attenuation in a narrow frequency range using dielectric elastomer stack actuators. In this controller, an estimation of the tremor frequency ensures suppression of only the tremor motion. The adaptive filter estimates the tremor torque, and a force controller for the dielectric elastomer tracks the specified force. Simulations show excellent tracking of the desired motion for slower voluntary motions and for slowly varying tremor amplitudes. Even though the controller has diminished tremor suppression in the presence of rapid changes in tremor amplitude, it still offers a significant improvement over the uncontrolled case. Altogether, this work demonstrates the potential for the use of dielectric elastomer actuators in a soft orthosis to suppress pathological tremor.