This paper conducted a study on the classification of basic hand movement from the two-channel surface Electromyogram (sEMG) data acquired from the upper limb. For classification, the scattering coefficients are obtained from sEMG signals by applying wavelet scattering transform with different quality (Q) factors. The statistical features namely mean value, summation, root mean square value, variance, maximum value, kurtosis, and skewness were computed from the scattering coefficients. The features were ranked using the ReliefF algorithm and classified using different machine learning algorithms to determine the most accurate classifier model. A comparative study of the performance of different classifiers in terms of accuracy has been done in this study. The proposed method has been applied to the sEMG data obtained from five subjects. The simulation results show that the average classification accuracy achieved by the proposed method is 93.4%.