People / Students / Class of 2025
My name is Tianyi Zhang, and I graduated from the University of Rochester with a Bachelor of Science in Data Science, concentrating on Computer Science, Statistics, and Mathematics. During my time at the University of Rochester, I gained valuable research experience as a Lab Assistant at the Audio Information Research Lab. Here, I leveraged advanced techniques such as discrete wavelet transforms and Mel-frequency cepstral coefficients for robust feature extraction and developed deep learning models integrating bidirectional LSTM networks and CNNs. My work involved applying statistical methods for precise event modeling and devising a dynamic programming approach for beat tracking, further enhancing my analytical skills. Additionally, I fine-tuned Llama 2 for a table game chatbot using Quantized Low-Rank Adaptation for efficiency and Retrieval-Augmented Generation for context-aware interactions, crafting the user interface using Gradio. I also engineered a Multi-Agent System using OpenAI's API, successfully enhancing music generation quality through innovative protocols, including a reflection system, surpassing standard outputs in complexity and creativity. In addition to my academic and research experiences, I have practical industry experience through two internships. As a Data Engineer Intern at the Agricultural Bank of China, I analyzed customer transactions and contributed to predictive analytics for customer churn. This summer, my role as a quantitative analyst at a Fund Management Company also allowed me to delve into performance attribution and factor exploration for FOF, further honing my data-driven problem-solving abilities. I am proficient in a variety of programming languages and tools, including Python, SQL, Java, MATLAB, C++, R, and Swift. Additionally, I have experience with data visualization and analysis tools such as Tableau and Gradio, as well as machine learning frameworks like TensorFlow, Keras, and PyTorch. I am skilled in designing and deploying machine learning models to solve real-world problems. My expertise includes feature engineering, model optimization, and end-to-end pipeline development. I am eager to join the MLDS program to deepen my knowledge and learn from distinguished faculty and talented peers. I believe that engaging with a diverse group of experts and fellow students will provide invaluable insights and foster collaborative learning. I am excited to contribute to and grow within this dynamic community, working together to achieve advancements in the field.
