Yihan Guo

Master's

Data Science

Graph Origin

- Extracted data, Transform data with persistent storage solutions - Managed cloud-based data, code and environment - Performed data cleaning, filtering and processing for AIGC model training - Trained machine learning models for automated data annotation - Designed and optimized data annotation work flows - Conducted model performance testing and identified optimal use cases and defects - Provided feedback through documentation to colleagues - Made tutorial presentations to other departments and clients - Developed distributed web scraping systems - Employed XPath, HTTP requests, and Selenium for data scraping - Project. Facial Expression Data Classification and Annotation Model - Responsible for designing a classifying and annotating system for 20,000 images of facial expression - Designed and developed a web-based user interface for efficient data visualization, annotation, and management - Experimented with linear regression model and deep face model - Explored research paper and code repository - Discovered a strong correlation between model performance and dataset - Improved model performance from 67% accuracy to 80% accuracy through image processing and filtering - Implemented SVM and CNN models to further enhance data annotation accuracy to 88 - Improved the universality of the classification method by introducing continuous variables for expression labeling and training linear regression models - ACADEMIC EXPERIENCE