I’m an engineer, computer scientist, and ML/data practitioner with a research background at FAU Erlangen-Nürnberg (Chair of Computer Science 6 / Evolutionary Data Management).
In recent years, my work has focused on large-scale time series integration and time series data mining — with methods developed in this context also proving useful in other domains, including interplanetary robotics.
For this research, I collaborate with Siemens Energy, and my initial project on data integration has expanded to large-scale data mining.
In my current role, I'm mostly translating open-ended problem statements into structured approaches, algorithmic designs, and implementable solutions.
Some of my prototypes have even left localhost and are now hosted as internal applications!
What I enjoy most is working closely with the domain experts who own the problem: I thoroughly enjoy learning their perspective, new domain knowledge, specific constraints, and consolidating their input into useful algorithms and data processing systems.
I appreciate interdisciplinary topics because they offer so much to learn, the opportunity to meet experts from other domains, and also leave room for me to contribute my own expertise.
I bring a fairly broad technical background across different application areas. I originally studied information technology (electrical engineering) with a focus on signal processing and have previously worked on topics including biomedical signal processing, algorithmic evaluation of automotive lighting systems, a short phase in computer vision (pose estimation), and biomedical applications.
Alongside my technical expertise, I have experiences with research project management, project acquisition, teaching on university level, and supervision of students.
I care about useful ML/data solutions — not just strong results on paper — and I enjoy taking ideas from exploration to implementation.
📌 I’ll soon be finishing my time at the university, and I am searching for new opportunities for turning data-heavy problems into reliable tools, workflows, and decisions — even if your problem is not primarily "time series".
A Python package for change point detection. (SST, IKA-SST, ESST, uLSIF, RuLSIF, KLIEP, FLUSS, FLOSS, and more)
After benefiting from open-source software countless times for many years, I wanted to do my part.
The package contains methods that I either used in my research, proposed, or contributed to. I wanted to make my research accessible and usable in other applications, so I consolidated the methods into a pip-installable Python package. I currently use the package in the core of the data mining applications that I'm building for my project partners.
Accelerating change point detection by orders of magnitude (SST, O(N³) -> O(NlogN)).
For this one, I had the opportunity to go down the absolutely fascinating rabbit hole of (randomized) linear algebra to solve an actual scaling problem in my research. Learning about all the great work in this area has been a lot of fun!
The paper is publicly available for everyone (open access).
How my research on time-series integration and change-point detection applies to interplanetary drilling robots. Update!
How can we visualize large time series collections and their complex relationships?
- 🏛️ FAU Profile: Lucas Weber (Chair page)
- 🎓 Google Scholar: Scholar profile
- 🆔 ORCID: 0000-0002-6877-6935
I have received many automated messages and have retracted my public email address. You can contact me by visiting my FAU profile, which includes my email address (see above), or by creating an issue in any of my public repositories.