Hand gesture is one example of a communication method for people with a hearing disorder. Currently, researchers have developed the stretchable sensing that is based on soft materials for supporting the devices of the conveys messages. Consequently, the real-time response and accuracy of the device are the major issue in the development process. However, the signal classification of soft sensors is still not widely used as a recognitive tool for human movement. In this current study, we applied the machine learning to understand and classify the human movement sign. This research presented the classification of the signal of the hand gesture language by using the reservoir computing algorithm. We expect that our reservoir computing model will classify the stretchable sensor sign. The results of the reservoir computing implementation classified the three signals of the hand languages. The reservoir computing’s performance showed that the weight input (W _in ) is the essential feature of the learning parameter. The significant parameters of reservoir computing support the optimizing performance of the classification.