The ability of triboelectric nanogenerators (TENGs) to sense physical, chemical, and physiological activities has been demonstrated. The data generated by TENG sensors encompass various parameters, including time, frequency, intensity, and acceleration. While this information can be used to effectively answer binary queries based on signal intensity, extracting additional intricate details requires an in-depth analysis of the collected TENG sensor data. Often, the amount of data amassed by these sensors surpasses the capability of efficient human analysis, necessitating the assistance of machine learning and deep learning approaches. Typically, supervised machine learning algorithms are employed for data processing, categorization, or identification. This paper provides a comprehensive review of recent advancements in machine learning for TENG sensors and highlights challenges to address in future research.