I would like to begin by thanking the generous donors who made this incredible opportunity possible for me through the Arts Student Employment Fund. My ARIA project, Estimating the Ordered Lorenz Curve and Gini Index for Machine Learning-Based Risk Prediction, supervised by Professor Archer Yang, focuses on improving model evaluation in insurance pricing and financial risk prediction. Building on recent research by Frees, Mario V. W眉thrich, Michel Denuit et al., the project examines how the ordered Lorenz curve and Gini index can better capture the relationship between predicted risks and actual outcomes, particularly under auto鈥慶alibration constraints. By exploring these tools within machine learning frameworks, this research aims to enhance model discrimination and calibration, contributing to more robust and interpretable risk assessment methods for complex insurance datasets.
As a third鈥憏ear student majoring in Statistics with a minor in Economics, I have been considering my future path 鈥� whether to continue with graduate studies or pursue a professional career after graduation. I sought an opportunity that would provide hands鈥憃n experience to help me make this decision. This was why I was particularly drawn to ARIA: it provided a unique chance to immerse myself in research, allowing me to experience the full process of academic inquiry beyond the classroom. This project was especially appealing because it integrates my two areas of study 鈥� Statistics and Economics 鈥� combining the statistical modeling and computational aspects of data science with the economic implications of risk prediction and insurance pricing.
My primary learning objective was to gain a systematic understanding of the research process. Unlike coursework, which is typically structured around lectures and exams, research involves identifying problems, formulating solutions, and iteratively refining approaches. I wanted to learn how to bridge the gap between theoretical knowledge and practical application, gaining skills in critical reading, analytical thinking, and academic communication.
One of the highlights of my internship was the intellectual growth I experienced. When I began, the field of actuarial science and insurance analytics was entirely new to me. Over the course of three months, I conducted an in鈥慸epth review of key literature, including foundational papers on the ordered Lorenz curve, Gini indices, and auto鈥慶alibration. Initially, even reading a single page could take hours, as I needed to look up terminology and underlying concepts. However, through consistent effort, I progressed from struggling to comprehend these papers to synthesizing their ideas into a comprehensive literature review. This deep engagement helped me develop a broad understanding of the field and its current challenges. Though ARIA has concluded, my research journey will likely continue, as I am eager to further advance this project and build on the work we have started during ARIA.
The most significant challenge I encountered was the steep learning curve of entering an unfamiliar domain. Early on, I often found myself reading without generating meaningful insights, simply grasping surface鈥憀evel content. I realized that true research requires not only understanding what the authors are saying but also critically reflecting on why their approaches work, where their limitations and how they might be extended. To overcome this, I adopted a deliberate approach: reading papers multiple times, discussing ideas with my supervisor, and gradually building connections between different concepts. This process, though slow, transformed my understanding and improved my ability to think independently about complex problems.
ARIA has had a profound impact on how I view my academic and career goals. By experiencing the realities of research, I now have a much clearer understanding of what day鈥憈o鈥慸ay research involves. This internship has strengthened my interest in pursuing further studies, potentially at the master鈥檚 or doctoral level, and has broadened my perspective on the intersections of statistics, economics, and risk analytics.
The financial support I received through the Arts Student Employment Fund was instrumental to my success in this program. With this funding covering most of my living expenses, I only needed to work around five hours per week at my part鈥憈ime job to cover the remaining costs, which allowed me to dedicate significantly more time and energy to my research. Without this support, I would have had to work many more hours to sustain myself, leaving far less time for academic engagement. The award not only eased my financial burden but also greatly enriched my overall ARIA experience by enabling me to fully immerse myself in the research process.
