Liu's journey began during his M.S. in Biostatistics program in 2020, where he expressed a keen interest in collaboration and a desire to contribute to impactful research.

Recognizing Liu's exceptional work and promising trajectory, Dr. Daniel Diaz, research assistant professor in the Division of Biostatistics, encouraged him to apply for the Institute for Data Science and Computing (IDSC) Fellows Program.

The IDSC strives to encourage a new generation of data science professionals to push beyond traditional discipline boundaries. IDSC collaborates with University academics to advise IDSC Fellows on how to best use the University's cutting-edge advanced computing facilities for their planned research projects.

“Liu's success in securing the fellowship is proof of his dedication and promising trajectory in ongoing research,” said Dr. Diaz.

Under Dr. Diaz’s guidance and in collaboration with Dr. Sunil Rao, professor emeritus, Liu began an innovative project titled “High-Dimensional Mode Hunting Using Pettiest Components Analysis.” This initiative sought to address the challenges presented by large dimensions in machine learning and statistics.

Published in the prestigious IEEE Transactions of Pattern Analysis and Machine Intelligence in April 2023, the research established the optimality of "pettiest components" for identifying clusters using tools from information theory. Liu served as the first author – a significant achievement in his academic journey.

With the support of the IDSC Fellowship and a steadfast commitment to impactful research, Liu, in collaboration with Dr. Diaz, is actively engaged in extending their previous work on clusters in networks alongside Dr. Kamal Premaratne from the School of Engineering. Building upon this foundation, Liu's current focus is on his project, “Spectral Graph Theory Methods Using Eigenpairs With Small Eigenvalues.” His research endeavors emphasize pattern recognition and clustering through advanced statistical and machine learning methods. "As an IDSC Fellow, I aim to explore the utilization of eigenvectors with the smallest eigenvalues in random graph models," he stated.

“Liu's exceptional work propelled him to become the only student transitioning from our M.S. program in Biostatistics to the Ph.D. program,” Dr. Diaz said.

Liu and Dr. Diaz’s recent paper on forecasting COVID-19 surges incorporating information technology tools was accepted for publication in PLOS One, highlighting Liu's ongoing commitment to advancing data analysis and making meaningful contributions to the field.

Written by Deycha Torres Hernández, published on January 31, 2024.