 This study proposes a novel fabric myoelectric armband in a hybrid serial classifier, HSC, algorithm to reduce the electrode shift problem associated with surface electromyogram pattern recognition, EMGPR. The armband is designed to minimize the electrode shift problem by reducing the number of electrodes used. Additionally, the HSC algorithm is developed to eliminate the need for retraining over multiple days. This algorithm uses only the first days EMG data to achieve high classification accuracy on subsequent days. The results show that the proposed armband and HSC algorithm can be used together to provide reliable and accurate EMGPR for long-term use. This article was authored by Hui Wang, Ping Gao Huang, Qing Han Su, and others.