Outstanding postdoc spotlight: Oskar Elek

Date
Oskar Elek
By: Melissa Weckerle

This is one of five profiles spotlighting an outstanding postdoc as part of Baskin Engineering’s celebration of National Postdoc Appreciation Week, September 20–24. Visit Baskin Engineering News and Voices of Baskin of Engineering for additional profiles of outstanding postdocs.

Oskar Elek joined Baskin Engineering motivated to find ways he could blend his technical background and artistic skills to conduct experimental, creative research. Before joining the UCSC Creative Coding Lab in 2019 under the direction of Computational Media Professor Angus Forbes, Elek received his Ph.D. degree in computer graphics at MPI Informatik Saarbrucken in Germany.

The first project he undertook at Baskin Engineering was applying the Monte Carlo Physarum Machine (MCPM) algorithm—a method to find hidden structure in data, originally inspired by efforts to understand the biological phenomena of slime mold growth and foraging behavior—to reconstruct the cosmic web, an astronomical term used to describe the entire structure of the universe. Since then, he’s been involved in other projects that employ the MCPM algorithm, including one to find new methods to improve machine learning language processing. Another aspect of Elek’s work that draws from his graphics background deals with data visualization. He creates 3D, high-resolution renderings of data to make research findings easier to understand.

“The idea of the algorithmically simulated slime mold and its digital visualization is broader than astronomy. It could connect multiple scientific domains, even language and other abstract structures, by studying their datasets with a single model,” said Elek.

With the creative freedom the lab brings, Elek has been able to collaborate on various interdisciplinary projects and take on new partnerships.

Most recently, he’s been awarded a grant from the Center for Research in Open Source Software (CROSS). Starting in October, he will develop an open source software for the MCMP algorithm he’s been applying to different scientific domain research. The goal is to make the system user-friendly and to welcome collaborations.