People / Students / Class of 2024
Yumin Zhang graduated from University of Wisconsin – Madison with degrees in Statistics and Psychology. She acquired her first taste of data science innovation through her work as a lab assistant in the Culture & Cognition Lab. To measure cultural differences in cognition processes, she collected, quantified, and analyzed verbatim responses using R. This experience showed her the value of performing quantitative analysis to reduce the subjectivity of the assessment process and kindled her interest in delivering insights through data storytelling. Since then, she has devoted herself to data science, aiming to tightly integrate the frontiers of the field with domain knowledge to uncover hidden insights. Yumin’s undergraduate study endows her with strong competencies in statistical theory, mathematical modeling, and computational skills. Working with high dimensional data, she implemented dimensionality reduction techniques and for the first time understood the trade-offs between model accuracy and model interpretability in practice. The following year, she worked with a professor to model changes in plant species richness over time. She visualized the geographic distributions of plant species, reduced noise by plant classification, applied oversampling using SMOTE to tackle data imbalance, refined and parameterized the Dirichlet-Multinomial model. By going through the entire data analysis workflow and using different statistical models to solve relevant problems, she honed her skills to work with real data and to operationalize theoretical concepts into measurable variables. Upon graduation, Yumin joined CHS Inc., a global agriculture company, as a full-time data scientist leading the digital transformation of financial reporting. To improve the reporting process, she centralized, democratized, and optimized data via SQL and Python to remove the need for non-value adding reconciliation between disparate reports, freeing up the FP&A department 60% of the time in data preparation. Throughout the process, she enjoyed technical challenges in tandem with facilitating continuous improvement. For instance, she monitored gateway performance, identified slow-performing queries, and refactored the code to optimize the database schema, which reduced memory usage by 30%. Such manifold aspects of the role not only solidified her hard and soft skills, but also showed her the importance of aligning data priorities with business goals. In the short term, Yumin aspires to polish her analytical skills and sharpen her industrial insight through the interdisciplinary coursework and practicum projects. Her long-run goal is to become an ethical data science professional, encouraging wider engagement with analytics among employees to seize more business opportunities.
