People / Students / Class of 2026
Yifei Hang graduated from the University of Washington, Seattle in June 2025 with a bachelor’s degree in Applied & Computational Mathematical Sciences and minors in Mathematics and Data Science. She maintained a record of academic excellence, earning Dean’s List recognition every quarter and graduating with a GPA of 3.95. Driven by her strong interest in data science and solid foundation in mathematics, Yifei developed extensive expertise in machine learning, data analysis, visualization, and statistical modeling, while gaining proficiency in programming languages including Python, SQL, R, and Java. Beyond academics, Yifei conducted a series of research projects in both the service and art industries. In her hotel reservation cancellation prediction project, she employed federated learning to address industry demands for privacy-preserving solutions, creating a model that protected data while maintaining strong predictive performance. In her pixel art generation project, she built a CycleGAN model to transform real-world landscapes into pixel-art style. Confronted with limited and low-quality training data, she designed data augmentation strategies that enhanced model robustness, achieving a 50% improvement over the benchmark. These projects strengthened her expertise in privacy-preserving techniques, generative models, and computer vision, while sharpening her problem-solving skills.
Through the Varanasi Quantitative Undergraduate Summer Internship Program at UW, Yifei applied her skills in a real-world context, exploring advanced machine learning methods to address remote-sensing data gaps. She developed and deployed a chlorophyll-a gap-filling pipeline that involved large-scale data cleaning and transformation, U-Net and Conv-LSTM model construction, and comprehensive visualization. This work deepened her understanding of how machine learning can transform oceanographic research. Building on this experience, her project at OceanHackWeek 2025 saw her play a leading role in developing a Deep-Residual-U-Net model to efficiently downscale sea surface height (SSH) data. Close collaboration with oceanography experts further honed her interdisciplinary communication skills. During her internship at COMNOVA Co., Ltd., Yifei contributed to the development of an intelligent OCR system tailored to extract client-specified information. She researched and evaluated OCR solutions for documents and images with a particular focus on table parsing and extraction. By annotating client data and fine-tuning models, she boosted accuracy on domain-specific documents, achieving a 30% improvement in layout detection and table structure recognition. She also integrated OCR tools into an intelligent document analysis framework, implementing data processing components to ensure full end-to-end functionality and system compatibility. Eager to continuously expand her expertise and apply data science to practical challenges, Yifei looks forward to joining the MLDS community, where she aims to leverage her skills to create impactful, real-world solutions.
