The objective of the proposed research is to revolutionize the field of Micro-Electro-Mechanical Systems (MEMS) by developing high-performance 4H-Silicon Carbide (4H-SiC) MEMS resonators and gyroscopes with machine learning-aided temperature compensation and calibration. The research specifically targets the longstanding challenges of temperature-induced frequency drift and material softening in 4H-SiC MEMS devices. By employing machine learning algorithms, the project seeks to dynamically model and correct temperature-related variations in real-time, thereby enhancing device stability and performance. The methodology involves a multi-phase approach that includes material characterization, device fabrication, and machine learning model development. Anticipated outcomes include a 4H-SiC MEMS resonator and gyroscope with significantly reduced temperature-induced frequency drift and a machine-learning model for real-time temperature compensation. This research has the potential to set new benchmarks in MEMS-based timekeeping and inertial sensing technologies, impacting various sectors such as telecommunications, aerospace, and healthcare.