<p>Capacitive sensors are widely used in industrial, biomedical, and consumer applications due to their high sensitivity and non-contact operation. However, temperature fluctuations can significantly degrade their accuracy by altering dielectric properties and sensor geometry. This study investigates the influence of temperature on capacitance and proposes a comprehensive compensation strategy to enhance measurement reliability. Analytical models incorporating temperature-dependent permittivity and thermal expansion effects are developed to predict capacitance variations. Simulations are conducted using both linear and polynomial regression, and the results are validated using machine learning techniques, such as Random Forest and Support Vector Regression. Design enhancements, including shielding, insulation, and low thermal expansion materials, are also evaluated. Results show that machine learning models outperform traditional compensation methods, reducing capacitance error by over 40% across a wide temperature range. The combination of insulation and shielding further improves thermal stability by up to 70%. These findings highlight the importance of integrating adaptive compensation algorithms with robust material design to ensure long-term sensor accuracy. The proposed approach enables more reliable capacitive sensing in environments with variable or extreme temperatures.</p>