Human hand gestures are an interactive and natural way of communicating and controlling, including their use in technology interaction, robotic manipulation, sign language, and so on. Correct hand gestures can be interpreted using several methods, including sensor-based techniques and machine-learning algorithms. This study presents a novel approach to the detection of human hand gestures using a combination of electromyography (EMG) and force myography (FMG) signals. The combination of EMG, which captures muscle electrical activity, and FMG, which measures the mechanical forces exerted by muscles, provides robust information and helps to classify gestures accurately. Custom data was recorded from 15 subjects to classify 15 hand gestures. XGBoost model achieved the highest performance, with a training accuracy of 93.70% and a testing accuracy of 80%. The proposed system shows great potential for use in prosthetics, human-computer interaction, and rehabilitation. The experimental results show that the combination of single-channel EMG - FMG signals is enough to classify multiple hand gestures with decent accuracy compared to multichannel EMG - FMG. Our research focuses on systems of cost-effective, compact sensory systems and the efficient use of machine learning to recognize complex gestures with precision.