Hello, I'm Ziyue Yin

A Data Science Student at Duke Kunshan University (DKU), passionate about AI, machine learning, and innovative solutions

About Me

I'm a passionate Data Science student pursuing a dual degree from Duke Kunshan University and Duke University (Class of 2026). My academic journey spans across data science, machine learning, and artificial intelligence.

With extensive research experience in speech enhancement, fake news detection, and large language models, I'm proficient in Python, machine learning frameworks, and data analysis. I'm currently maintaining a 3.806/4.0 GPA and have been recognized on the Dean's List.

Beyond academics, I'm actively involved in entrepreneurship and innovation initiatives, leading teams and coordinating high-impact events. I'm passionate about bridging technology and real-world applications.

3.81

GPA / 4.0

6+

Research Projects

2026

Graduation Year

Skills & Technologies

Programming Languages

Python Java Shell MATLAB JavaScript

AI & Machine Learning

TensorFlow PyTorch Llama3-8B RAG NLP Computer Vision

Tools & Technologies

AGI (ChatGPT/DeepSeek) LaTeX Adobe Illustrator KEGG Trinity BLAST

Languages

Chinese (Native) English (Proficient) Spanish (Beginner)

Education

Duke Kunshan University & Duke University

Dual Degree Program

2022 - 2026 (Expected)

B.S. in Data Science

Duke Kunshan University, Kunshan, China

B.S. in Interdisciplinary Studies: Data Science

Duke University, Durham, U.S.

GPA: 3.806/4.0
Academic Honor: Dean's List (Fall 2022, Spring 2023, Spring 2025)

Featured Projects

TrustNet: Fake News Detection

Leading a team to develop an AI model integrating text analysis for enhanced fake news detection, processing 10,000+ social media articles with optimized accuracy using TensorFlow & PyTorch.

Python TensorFlow PyTorch NLP Machine Learning

Speech Enhancement with Audio-Visual Models

Research on real-time, low-latency speech conversion pipeline using compressed generative models for whisper and electro-laryngeal speech enhancement, exploring large generative models for zero-shot naturalness.

Python Audio Processing Deep Learning Computer Vision Signal Processing

Large Language Models for Q&A

Developed a Retrieval Augmented Generation (RAG) pipeline integrating Llama3-8B for question-answering, achieving 21.4% improvement in Hit Rate and 12.9% in Context Recall through optimized hyperparameters and fine-tuned embedding models.

Python Llama3-8B RAG NLP Fine-tuning

Multi-omics Data Analysis

Reconstructed genome-scale metabolic network for water-bloom cyanobacteria using KEGG reactions, processed large-scale transcriptomics data with Trinity, RSEM, and BLAST for identifying temporal metabolic shifts.

Python Bioinformatics KEGG Trinity BLAST

Dots Connected

Interests

Badminton

Love the thrill of competitive matches and the feeling of staying fit!

Table Tennis

Quick reflexes and strategic thinking - perfect for clearing my mind!

Singing

There's something magical about letting emotions flow through music

Piano

My go-to stress reliever - nothing beats the feeling of creating melodies

Blog Coming Soon!

Working on a space to share the little moments, big ideas, and everything in between. Coming soon!

Life Stories
Photo Moments
Random Thoughts
View Progress

Get In Touch

I'm always interested in new opportunities and collaborations. Feel free to reach out if you'd like to work together!