Talk 1 on Human-AI Cooperation for Fairness Elicitation
Talk at Fair and Explainable Decision Making (FED) lab, University of Massachusetts Amherst, Amherst, Massachusetts
Chang Zeng's CV
Education
- B.S. in University of Massachusetts Amherst, 2018-2022
- Major in Computer Science
- Major in Environmental Science
- M.S. in University of Massachusetts Amherst, 2022-2024 (expected)
- Major in Computer Science
Awards
- Bay State Fellowship at the University of Massachusetts Amherst.
- Dean’s List recipient for the following semesters: Fall 2018, Spring 2019, Fall 2019, Spring 2020, Fall 2020, Spring 2021, Fall 2021, and Spring 2022 at the University of Massachusetts Amherst.
Research Experience
- Human-AI Cooperation for Personalized Fairness Optimization - University of Massachusetts Amherst - 2022-Present
- Supervisor: Cyrus Cousins and Yair Zick
- Contribution:
- Analyzing and Utilizing Power Mean Fairness Concepts to elicit $\varepsilon$-approximate Human Cardinal Fairness Concept.
Work Experience
- Software Engineer Intern - X-Camp - 2023-Present
- Supervisor: Yuan Xu
- Contribution:
- Zoom Integration with Education Platform
- Detect & Remove Duplication using NLP Technique
- Research Assistant - University of Massachusetts Amherst - 2022-Present
- Supervisor: Cyrus Cousins, Yair Zick
- Contribution:
- Research on the Human-AI Cooperation for Fairness Elicitation
- Teaching Assistant - University of Massachusetts Amherst - 2022-Present
- Supervisor: Parviz Kermani
- Contribution:
- Host weekly office hour and be accessible for students in many forms, such as responding piazza in short time
- Prepare and design the homework, exam, and challenge to help student have a better understanding of the concept
- Grade homework and exam
Project Experience
- Autoencoder Optimization - Amherst, MA - 2022-2023
- Software Engineer - Back-End (Python, Deep Learning, MLP, CNN, RNN)
- Contribution:
- Developed an autoencoder with a backpropagation mechanism to effectively denoise individual heartbeats from electrocardiogram signals, enhancing signal quality and accuracy.
- Performed a comprehensive comparison between linear and non-linear structures, including CNN and RNN, to evaluate their performance in optimizing the denoising process.
- Employed academic research methodologies to methodically fine-tune autoencoder model parameters, incorporating strategies like scheduled learning rate adjustments and varied layer structures, resulting in elevated denoising capabilities and heightened accuracy.
- Wuhuu Information Sharing Platform - Amherst, MA - 2022-2023
- Software Engineer - Back-End (Python, Crawler, AWS S3)
- Contribution:
- Developed customized crawlers utilizing Python to systematically gather data from social platforms, focusing on user engagement trends, content popularity, and sentiment analysis.
- Implemented a robust storage infrastructure using AWS S3 to securely store and retrieve shared files, while optimizing data retrieval and minimizing latency.
- FFIV Ultimate Simulator - Amherst, MA - 2022-2023
- Lead Developer - Full-Stack (Unity, C#, WPF, UI/UX, XAML)
- Contribution:
- Spearheaded the design and construction of a Unity game simulator, employing C# to enable players to customize and personalize their gameplay tactics allowing them to practice and refine their strategies.
- Orchestrated the creation of a tailor-made launcher for the simulator, leveraging the WPF framework to deliver a user-friendly interface that streamlined access to game features and settings.
- Eye Activity Tracking - Amherst, MA - 2022-2022
- Software Engineer - Full-Stack (Python, Machine Learning, OpenCV, CUDA)
- Contribution:
- Employed various machine learning models, including the BERT transformer (TensorFlow), along with advanced techniques such as random forests and stratified k-fold, to accurately classify eye behavior patterns.
- Utilized OpenCV libraries to capture and process live camera feed in real-time, extracting relevant features from the eye images, such as pupil dilation and eye movement, to analyze and determine the user’s eye activity.
Skills
- Programming Languages:
- Python, Go, C#, HTML, CSS, XAML, TypeScript, JavaScript, C++
- Library:
- TensorFlow, PyTorch, NumPy, Pandas, Flask, FastAPI
- Framework:
- Kubernetes, Docker, OpenCV, React, Node.js, PostgreSQL
- Tools:
- Git, Github, Gitlab, AWS S3, OSS, Bash, Conda
- Operating System:
- Unix/Linux, OSX/IOS, Windows
Talks
Talk 2 on Human-AI Cooperation for Fairness Elicitation
Talk at Fair and Explainable Decision Making (FED) lab, University of Massachusetts Amherst, Amherst, Massachusetts