We are pleased to announce that David Kepplinger has won the Department’s 2019–20 Data Science Award.
David is finishing his PhD degree in Statistics under the supervision of Prof. Gabriela Cohen Freue, with a focus on robust regularized regression of high dimensional data. His work on a Penalized Elastic Net S-Estimator (PENSE) is particularly noteworthy. Here, David integrates statistical methodology, sophisticated optimization algorithms, computational development, implementation, and data analysis, key components of Data Science. David’s code is available in the R package pense. He has applied his methods to the analysis of proteomics data, identifying new potentially relevant biomarkers of cardiac allograft vasculopathy that are not found with alternative methodologies. This approach improved the classification of independent test samples. This work is featured in an article in the Annals of Applied Statistics (2019).
In addition to his work on PENSE, David contributed to the automated greenhouse management platform of Ecoation Innovative Solutions Inc. through an NSERC Engage internship. His work on a project to detect plant stress by an automated system was so well-received that his contract was extended so that he could build a cloud-based machine learning platform to manage and analyze the massive amounts of data collected daily in greenhouses in several parts of the globe.
David has also contributed to computing tools for the analysis of protein data. He has developed an algorithm to link protein groups created from high dimensional mass-spectrometry data (PLoS ONE paper and Bioconductor package ‘pgca’). Another notable example of his computational endeavors is the development of the Shiny application to visualize a high-volume mRNA and protein data. David developed the Shiny app to augment the Nature paper written by Gabriela Cohen Freue. The app makes it easy for researchers to visualize the data analyzed in the paper and to showcase common analysis flaws that were discussed.
Last but not least, David has greatly contributed to the training of researchers and students in Data Science skills. He has been an outstanding mentor to his fellow students. As a teaching assistant in data oriented courses, he has blended Data Science pedagogies and tools (e.g., RMarkdown reports, Jupyter Notebooks, GitHub repositories, and tidyverse code) into traditional statistical curricula. David has also played an essential role in the R workshops offered by the Statistics Department’s Applied Statistics and Data Science Group.
For more information on this award, please see https://www.stat.ubc.ca/data-science-award.