Machine Learning Data Science

Zhuowen Ye

Houston, TX — *************@*******.***— 909-***-****

EDUCATION

Rice University Master of Data Science GPA:3.71/4.0 2023-2024 UNC-Chapel Hill BSc in Statistics & Analytics GPA:3.68/4.0 2021-2023 SKILLS & COURSEWORK

Programming: Python, R, SQL, MATLAB, HTML/CSS/JS, Git, Linux, AMPL, AWS Tools: Pytorch, TensorFlow, Keras, Spark, Hadoop, Pandas, Numpy, Matplotlib, Tableau Relevant Coursework: Calculus, Linear Algebra, Discrete Math, Data Management & Data Science, Analyt- ical Geometry, Vector Analysis, Data Structure, Database, Optimization for ML and NN, Probability, Machine Learning, Dynamic Decision Analytics, Methods Data Analysis, Deep Learning, Stochastic Modeling PROFESSIONAL EXPERIENCE

Fosun Pharma - Aitrox Shanghai, China

Machine Learning Engineer Intern May 2024 - Aug 2024

• Automated a de-identification algorithm using YOLOv7 for processing over 10,000 images, videos, and DICOM ultrasound files, ensuring patient privacy by identifying and obscuring personal information.

• Detected and blacked out patient personal information on ultrasound files, enhancing data security, ensuring compliance with privacy regulations, and protecting patient confidentiality.

• Developed a fully autonomous robotic ultrasound system for thyroid scanning, leveraging imitation learn- ing and SLAM to detect malignant nodules without human assistance, reducing scan and diagnosis time by 50%. Collected and preprocessed expert demonstration data, trained models via behavior cloning and inverse reinforcement learning, and implemented Visual SLAM for real-time environment mapping and path planning.

Texas Children’s Hospital Houston, TX

Advanced Machine Learning for Pediatric Cardiomyopathy Diagnosis Jan 2024 - April 2024

• Enhanced preprocessing of Doppler spectrograms and strain mappings through image cropping, and advanced ECG analysis, significantly optimizing data quality for machine learning model inputs.

• Developed separate machine learning models for ECG images, tabular data, and strain mapping images, employing logistic regression, tree-based methods, and CNN architectures, followed by comprehen- sive model evaluation using ROC-AUC and stability testing, setting a foundation for future integration into a unified model for binary classification tasks.

• Utilized Grad-CAM for interpretative analysis, producing heatmaps to illuminate CNNs’ decision- making, aligning machine-identified critical features with clinical diagnostic standards to support doctors in validating and refining diagnosis protocols.

RESEARCH EXPERIENCE

UNC-Chapel Hill, The VisuaLab Chapel Hill, NC

Research Assistant, Modeling High-Level Visual Comprehension Aug 2022 - Nov. 2022 Supervised by Dr. Quadri

• Spearheaded a qualitative analysis project, utilizing Convolutional Neural Networks to interpret com- plex human-visualization interaction patterns and enhance intuitive data presentation strategies.

• Developed a Transformer-based model to automatically parse and summarize a corpus of over 1,400 high- level empirical visual comprehension datasets, leveraging NLP techniques for efficient knowledge synthesis.

• Conducted rigorous statistical analysis on comprehensive visual data, deriving over 100 key metrics to refine model accuracy and deepen insights into visual cognition, laying the groundwork for sophisticated analytical tool development.

PROJECTS

Kaggle Competition: Neuron Synapse Prediction Jan. 2024 - Apr. 2024

• Optimized Data Processing, enhanced preprocessing techniques including StandardScaler for normaliza- tion and target encoding for categorical features, improving data readiness for predictive modeling.

• Developed Predictive Models, built diverse machine learning models such as Logistic Regression, Gra- dient Boosting, XGBoost, and neural networks, tailored for class imbalance with SMOTE.

• Conducted comprehensive evaluations including decision threshold optimization, achieving a balanced ac- curacy of 0.78614 and ranking 17th.

Contact this candidate