Research
This page documents one of my research learning experiences and highlights my current interest in applying machine learning, particularly large language models. I aim to explore how these tools can be made more efficient and impactful in real-world economic contexts.
Reflection on Directed Reading Program
—Exploration of LLMs
The Directed Research Program provides me with an opportunity to deepen my understanding of large language models (LLMs) and their growing influence across disciplines. Engaging in discussions with a PhD instructor and reading academic papers gave me firsthand exposure to the complexity and elegance of transformer architectures and attention mechanisms—the very foundations of modern LLMs.
One of the most valuable aspects of the program was learning how these models process unstructured data. Previously, I had only considered data analysis primarily in structured formats, like spreadsheets and databases. However, LLMs opened up an entirely new dimension: the ability to parse, interpret, and generate insights from vast volumes of human language. This shift expanded my thinking on what is possible in data science, especially in fields that rely heavily on qualitative information, such as finance and economics. I began to see these fields not only as domains driven by numerical models but also as rich sources of language-based data, such as policy announcements and market commentary, that can offer powerful insights when analyzed through the lens of LLMs. This interdisciplinary potential has solidified my motivation to pursue further research in this space, ideally at the intersection of AI and social sciences.
Overall, the Directed Research Program taught me how to read technical papers more effectively and think critically about innovation. It showed me that curiosity, when coupled with mentorship, can lead to genuinely transformative learning.
Research Interest
My current research interests lie at the intersection of advanced machine learning and financial data analysis. What draws me to this field is the potential of emerging technologies, especially large language models, to make sense of complex, unstructured information and drive more informed decision-making. This curiosity began with experiences applying mathematical tools to real-world problems and has grown through collaborative research and data-driven projects.