Author: Jaxon Preissinger | Major: Data Science Social Data Analytics | Semester: Summer 2025
Throughout the Summer 2025 term and the upcoming Fall 2025 term, I have been and will continue to be working with my honors mentor, Dr. Rossetti, in creating a systematic method of fitting mixture distributions within the programming language, Kotlin. Mixture distributions analyze datasets with multiple subpopulations and find the best solution to the distribution problem presented by this data. In simpler terms, this means that when you have a set of data, you can separate it into multiple groups based on the spaces between each data point, often seen as humps in a graph, and then assign each group a given “fit” that best matches.
Across many fields of interest, mixture distributions play an important role due to their ability to model multimodal data. For example, they help model stock return distributions in finance, identifying multiple market behaviors within a single dataset. For the machine learning field, mixture distributions form the basis of clustering algorithms such as Gaussian Mixture Models, which provide some of the framework for machine learning models. Additionally, in the epidemiology field, mixture distributions are being used to discover disease subtypes and assess infection spread; this was incredibly helpful during the COVID-19 outbreak. Furthermore, in cancer research, mixture models are used to identify distinct tumor subtypes, which is vital for developing targeted cancer therapies and advancing medical precision. However, despite the impact that mixture distribution modeling can have, no library within Kotlin can perform this task. So, to explain how I found this issue, the short answer is: Dr. Rossetti did, and I stumbled into this fantastic opportunity. In meeting Dr. Rossetti for the first time, I wanted to apply for a job helping the Data Science department. Coincidentally, this was on the same day as the Engineering Career Fair, so maybe it was my suit, but after Dr. Rossetti told me the position “has already been filled,” he offered me an opportunity to help with an upcoming research project with him. With a swift decision, I accepted.
In addition, throughout this process, Dr. Rossetti has served as a mentor and expert, not only helping me understand the subject matter but also encouraging me to return with questions, answers, and problems to overcome together. Furthermore, Dr. Rossetti has been the sole reason for my ability to progress in this subject matter, as his encouragement has kept me grounded. His research expertise has helped me focus on one task at a time and avoid burning out.
Furthermore, this research has not come without challenge; to accomplish ample work over the summer session presented difficulties of its own. On my end, I struggled to keep up with this research over the summer, as I needed to travel domestically and abroad to see family and take vacations. Being away for several weeks at a time made it essential to create a routine to help maintain a structure. With remote meetings, Dr. Rossetti and I were able to set and accomplish weekly tasks to keep up with the workflow and continue making progress towards our end goal.
Moving forward into the next semester, our weekly structure is set, and our goal is straightforward: to finalize the mixture distribution modeling. Progress will continue, and our research will be completed on time. Also, I want to note how incredibly thankful I am to have received the Honors College Research Grant, which has allowed me to research and develop something that can help others in meaningful ways.