Yujie (Stefanie) Li

802-***-**** adkjyh@r.postjobfree.com LinkedIn: in/yujie-li/

SUMMARY

Data Scientist with quantitative research background. Over six years of experience applying data science and analytics techniques to solve practical problems. Expert in analyzing structured/unstructured big data and building statistical/predictive models to deliver deep insights and data-driven solutions. SKILLS

Programming: Python (Scikit-Learn, Numpy, Pandas, Seaborn, SciPy), SQL, Tableau, PySpark, AWS, GCP, Git, Java, XML, MATLAB, C

Machine Learning and Data Science: Data Mining/Visualization Statistical/Predictive Modeling, Feature Engineering, Machine Learning (Linear/Logistic Regression, SVM, Random Forest, XGBoost, Time Series, Clustering), Bayesians, DOE (A/B Testing), Deep Learning (CNN, RNN, LSTM), Natural Language Processing (LSA, NMF, LDA).

WORK EXPERIENCE

Research Associate • Oak Ridge National Laboratory 03/2019 - present Oak Ridge, TN

• Developed the core algorithm for a workflow with Python and MATLAB to analyze and visualize large scale imaging data and save analysis time by ~40%.

• Built a classification model to predict the phase degradation of irradiated cementitious materials, engineered features and implemented logistic regression, kNN, random forest, neural network and convolutional neural networks (CNN) models for X-ray fluorescence data to generate phase distributions with a precision of 0.9.

Data Scientist Fellow • Techlent 08/2020 - present Oak Ridge, TN

• Established an inventory forecasting platform to track historical sales data and provide forecasting of opening week sales demand. Cleaned and processed big sales dataset, applied NLP and feature engineering and optimized an XGBoost model to predict demand, wrapped the model as an API with Flask and deployed the system to Google Cloud Platform (GCP). Research Assistant • University of Vermont 08/ 2013-12/ 2018 Burlington, VT

• Developed a classification model to predict air quality, compared different machine learning models including SVM, logistic regression, random forest, XGBoost with bagging and stacking to predict PM2.5 level in different cities, and increased the AUC scores from 0.6 to 0.93.

• Conducted parametric studies to optimize the modified asphalt compositions. Built a pipeline to predict the material compositions with different recipes and implemented statistical methods to comprehensively evaluate effects of different factors on the microstructures and mechanical properties of bituminous materials.

EDUCATION

M.S. in Computer Science Georgia Institute of Technology, Atlanta, GA 2021 Ph.D. in Engineering University of Vermont, Burlington, VT 2018 B.S./M.S. in Materials Science Huazhong University of Science and Technology, China 2010/2013