Muscle fatigue is a prevalent issue that impacts an individual's performance and health in the fields of medicine and rehabilitation. In order to enhance our understanding and analysis of muscle exhaustion, the utilization of Electromyography (EMG) signals has gained prominence. This research aims to analyze muscular exhaustion by employing features derived from surface electromyography (sEMG) data. Data was collected, and a comprehensive analysis was conducted by recording electromyography (EMG) signals from the experimental group during exercise activities. The examination of sEMG signals reveals that factors in both the temporal and frequency domains may provide crucial information regarding the extent of muscle exhaustion. Describing the utilized models or approaches in the research. Two experimental datasets are employed in this investigation to assess muscle fatigue. The algorithm described above extracts and filters a total of fourteen features, of which eight are in the time domain, and six are in the frequency domain. In conjunction with these, the minimum Redundancy - Maximum Relevance (mRMR) feature selection method is implemented. Support Vector Machine (SVM), Linear Discriminant Analysis (LDA), and K-Nearest Neighbors (KNN) are the algorithms implemented in this paper. The research findings indicate that the KNN classification model yields the most favorable outcomes, achieving a specific F1-score of 95.12%. The characteristics of the classification model mean absolute value, skewness, and mean are thoroughly assessed. KNN demonstrates effective classification of muscle fatigue. The results obtained from this research provide valuable insights for decision-making and make a significant contribution to the improvement of performance and recuperation from muscle fatigue. Considering the potential of EMG signals for analyzing muscle fatigue, forthcoming investigations may focus on the practical implementation and integration of this technique within the domains of medicine and rehabilitation.