About Chin-Hsuan(Shawn) Sun

Hello, I am Chin-Hsuan (Shawn) Sun, as a Data Scientist at TSMC, I bring a unique blend of skills as a Full Stack Engineer and Infrastructure Engineer, specializing in DevOps and Site Reliability Engineering (SRE). My expertise lies in building robust, scalable systems that drive data-driven decisions and optimize operations. With a passion for continuous learning and innovation, I thrive on solving complex challenges at the intersection of data science and engineering, contributing to cutting-edge advancements in the semiconductor industry.

Was a summer intern in AI algorithm development at Airoha Technology, I focused on enhancing the accuracy of GNSS and dead reckoning for electric scooters. Traditional navigation systems using GNSS faced challenges due to signal obstruction from tall buildings. To address this, I explored the feasibility of integrating machine learning techniques, including LSTM and decision trees, to improve navigation reliability under these conditions.

I worked as a research intern at the National Institute of Informatics in Japan, where I developed a surrogate model for optimizing delivery robots. Collaborating with Panasonic, our work was published at an IEEE conference. You can view the publication here.

Graduating from the Graduate Institute of Biomedical Electronics and Bioinformatics at National Taiwan University, holding a Master’s degree. Throughout my academic journey, I have delved into the intersection of software engineering and biomedical optics, a field that captivates my passion for innovation and problem-solving.

My master’s thesis, titled “Accelerating Monte Carlo Simulation through Surrogate Models for Quantifying the Internal Jugular Vein using Artificial Neural Networks,” reflects my commitment to advancing research at the cutting edge of technology and biomedical optics. In this project, I explored the integration of computational techniques, leveraging surrogate models and artificial neural networks to enhance the efficiency of Monte Carlo simulations. Then, I built a prediction model by utilizing the surrogate model to generate a bunch of diffuse reflectance spectra as dataset for training. Therefore, it shows the ability to quantify the blood oxygen saturation change of internal jugular vein. This work not only showcases my technical proficiency but also underscores my dedication to addressing challenges in the realm of biomedical optics.