One of the bio-signals used to power a prosthetic hand and enhance the quality of life for amputees is myoelectric signal produced by the muscles. Furthermore, surface electromyography is the most promising technique for obtaining muscle activity; however, an effective pattern recognition system is required. This research presents a method based on EMG signals for simultaneous detection of wrist motions and hand gestures. In this method, the first signal is acquired using Myo armband with 36 intact subjects and seven hand motions have been recorded. Following signal preprocessing, ten time-domain features are retrieved, and the chosen features are then categorized using SVM, DT, RF, and XGBoost. The grid search technique is used to fine-tune the hyperparameters, and the results show which hyperparameters are optimal for the input to the classifier. Based on the experimental results, random forest was able to obtain 84.2 ± 2.7% classification accuracy. The suggested technique has been tried on a 12-generation Intel Core i5 1.30 GHz CPU equipped with 16 GB DDR4 RAM.