Investigation of Resonance Characteristics of More Than 10 000 Metal‐Coated Silicon MEMS Cantilevers Using Machine Learning

This work presents a machine learning‐driven framework for the systematic analysis of silicon micro‐cantilevers based on a large experimental dataset comprising more than 10 000 structures. By modeling resonance frequency and quality factor across multiple vibrational modes, device designs with surface coatings, the proposed approach reveals geometry‐ and material‐dependent trends that are difficult to capture using conventional analytical or numerical methods of MEMS resonators. This is the first study to leverage machine learning on such a large experimental dataset with different types of metallic surface coatings to investigate their resonance behavior. The results demonstrate that ensemble‐based models can achieve high predictive accuracy, with resonance frequency errors approaching experimental uncertainty and consistent performance across surface types. Overall, the proposed experimental validation and design optimization approach establishes a scalable and physics‐informed reference platform for MEMS resonator prediction, design optimization, and experimental validation.