This study investigates the impact of Tai-Chi practice on human gait and muscle activation patterns using deep learning-based classification. Tai-Chi, a traditional Chinese martial art, is known for improving balance, coordination, and lower-limb strength, but its quantifiable effects on force-sensitive resistor (FSR) gait biomechanics signals and electromyography (EMG) signals remain under explored. In this work, FSR gait and EMG signal database are employed which is publicly available via PhysioNet. There are three groups: control (non-practitioners), Tai-Chi practitioners, and masters (experienced practitioners). Features were extracted from segmented gait cycles and EMG signals, followed by classification using a deep multi-layer perceptron (MLP) model. The proposed model was trained and evaluated to distinguish between the three groups, providing insights into how Tai-Chi influences gait dynamics and neuromuscular control. Experimental results demonstrate that Tai-Chi practitioners exhibit distinct gait patterns compared to non-practitioners, with Masters showing more refined motor control characteristics. The findings suggest that Tai-Chi training induces measurable biomechanical adaptations, which can have implications for rehabilitation and balance training applications.