Machine Learning Software Engineer

Tracy Thao Duong

Software Engineer

********@*****.*** 408-***-**** github.com/tpduong linkedin.com/in/tracyduong EDUCATION

Georgia Institute of Technology - Master of Science, Computer Science Machine Learning specialization, GPA: 3.8 2020 - 2023 University of California, Davis - Bachelor of Science, Computer Science Davis, California

PROJECTS

Indiana Drones

Developed a Simultaneous Localization and Mapping (SLAM) module to track drone location using noisy sensor measurements and created an obstacle avoidance system to navigate a forest and search for treasure. Heart Disease

Analyzed Kaggle patient data to identify key heart disease risk factors, including age, cholesterol levels, and blood pressure, and performed feature selection.

Evaluated machine learning models (logistic regression, decision trees, neural networks, support vector machine, boosting) with hyperparameter tuning and conducted learning curve analysis to optimize prediction accuracy. Applied k-means and Expectation Maximization (EM) clustering algorithms along with dimension reduction techniques and evaluated performance using metrics like Adjusted Rand Score, Mutual Information, and V-measure. Improved clustering performance for heart disease data and confirmed the effectiveness of using Neural Network models on reduced-dimension data for better computation efficiency without significant accuracy loss. Achieved high accuracy in heart disease prediction, comparing training scores to cross-validation scores and assessing accuracy as a function of the number of training examples, and highlighted cholesterol and age as influential factors. SKILLS

Data: MySQL, MongoDB, Vertica

Frameworks/Libraries: AngularJS, Node.js, NumPy, Pandas, PyTorch, ReactJS, Redux, Scikit-Learn Machine Learning Algorithms: Decision Trees, K-Means Clustering, K-Nearest Neighbors, Linear Regression, Monte Carlo Methods, Naive Bayes, Neural Networks, Principal Component Analysis (PCA), Proportional-Integral-Derivative

(PID) Control, Q-Learning, Random Forest, Simultaneous Localization and Mapping (SLAM), Support Vector Machines, Temporal Difference (TD) Learning.

Programming Languages: HTML, Java, JavaScript, JSON, Python, SQL WORK EXPERIENCE

MachineZone - Software Engineer

Palo Alto, CA 2016 - 2018

Implemented Machine Zone's Marketing Platform Web Portal, a comprehensive media-buying platform for planning, creating, buying, optimizing, measuring, and reporting media around custom key performance indicators, leading to a revenue increase of over $1 million/day.

Analyzed extensive data from multiple sources, including internal game metrics, player behavior, purchases, advertising- partner demographics, device usage, and URL-tracking information, to target high-value players and enhance game monetization and marketing effectiveness.

Led full-stack development with Vertica DB, MySQL, Java with Spring, and frontends using Angular 1 and ReactJS/Redux, spearheading the platform's upgrade from Angular 1 to ReactJS/Redux to boost performance and code efficiency. Collaborated with product managers, data scientists, and marketers to develop and implement advanced fraud-detection tools, enhancing security measures and saving over $1 million in potential losses. Established and enforced best coding practices within the team, fostering a culture of quality and efficiency through code reviews, code analysis, and unit testing.

Contact this candidate