The EMG signal can be measured through the electrode during the contraction of the muscle, and various studies are currently being conducted to use it in engineering. In this paper, to classify human hand movements, a pattern recognition algorithm is constructed using the measured EMG signals of the lower arm. The EMG signals of the lower arm are measured and analyzed, and the features are extracted based on the EMG signals. Then, pattern recognition is performed by applying the artificial neural network algorithm. 8 channels of EMG signals are measured. Based on this, it is confirmed that the five hand movements can be classified into 98% accuracy.