Menu Open

Wed Apr 2 2025 - Fri Apr 25 Past

Wed - Fri 2 Apr 2025 - 25 Apr 2025Past

Seminar - Statistical and Machine Learning Methods

Training Hybrid (University of Duisburg-Essen)

Seminar Machine Learning Statistical learning By DKZ.2R

Course Description and Goals:

Statistical learning is a field that teaches students how to analyze and interpret data by applying statistical methods and machine learning algorithms to uncover patterns, make predictions, and gain insights from data. The syllabus includes:

• Statistical and machine learning methods, including linear and polynomial regression, logistic regression, and linear discriminant analysis.

• Model validation techniques such as cross-validation and bootstrap, model selection, and regularization methods (ridge and lasso).

• Nonlinear models, splines, and generalized additive models.

• Tree-based methods, including random forests and boosting.

• Support-vector machines and an introduction to causal inference.

• Unsupervised learning methods such as principal components analysis and clustering (k-means and hierarchical).

Examination Format: Report and Presentation (obligatory to receive participaation certificate)

General Information University: University of Duisburg-Essen

• Learning Platform: Moodle (Link: https://lehre.moodle.uni-due.de/course/view.php?id=1986)

• Hybrid Format: Yes, the course includes both in-person and online sessions via Zoom

• Study Program and Level: Master’s/PhD students

• Weekly Hours: The event consists of three block sessions (lecture with tutorial sessions) with two additional dates for tutorials:

• First session: 02.04. (9:00–14:00)

• Following sessions: 03.04. & 04.04. (9:00–17:00)

• Online session: 07.04. (starts at 10:00; end time not specified)

• Final lecture: 25.04. (9:00–14:00)

• Language: German or English (depending on student preference)

Further information, including locations, and Zoom links, can be found on our homepage: Statistical Learning SS 25

Tags:

How To: Open Science

A. Mielke
Data , RDM , Open Science

Tired of Recreating someone else’s work? - How Open Science can accelerate research and overcome reinvention

Have you ever found papers on algorithms but their implementation is missing? Found an interesting analysis but there is no way to check the results, as you don’t have access to the data they were derived from? Ever thought you had a great idea for a project, just to find out a year later that you are not the only research group following that specific idea? Not having access to other people’s code, data, metrics or even their plans for research projects often leads to unnecessary delays and scientific redundancies. There is an easy solution to overcome (almost) all of these issues. It’s called Open Science! What is Open Science? The UNESCO defines Open Science as a construct of “movements and practices aiming to make multilingual scientific knowledge openly available, accessible and reusable for everyone, to increase scientific collaborations and sharing of information for the benefits of science and society, and to open the processes of scientific knowledge creation, evaluation and communication to societal actors […]”. To ensure that everyone has access to scientific knowledge and infrastructure, Open Science focuses on four main concepts.

Data Competence College

The first Data Competence College will be hosted from March 26th to 28th at the IT center of RWTH Aachen. Based on the concept of the Wissenschaftskolleg in Berlin or the Institute of Advanced Studies in Princeton, we have invited four individuals with high data competence from different scientific fields (“Data Experts”) as part of the data competence college.

As of now, Prof. Sebastian Houben from Hochschule Bonn-Rhein-Sieg (specialist in AI and autonomous systems), and Dr. Moritz Wolter from University of Bonn (expert in high performance computing and machine learning) have already confirmed their participation.

FDM-Werkstatt - Into the RDM-Toolbox!

A. Janz
Event , RDM

The Center of Data Litercacy (German: “Zentrum für Datenkompetenz”) DKZ.2R was officially launched mid November 2023. Already a month later we joined forces with fdm.nrw to organise the very first DKZ.2R-event (find the call for participation here). The “FDM-Werkstatt – Into the RDM-Toolbox” took place from March 18 to 20, 2024 at the IT Center of RWTH Aachen University. In total, 50 participants from all over Germany took part in the workshops. Many of them brought their own topics of interest with them and presented it in one of the 12 sessions. The contents of the sessions ranged from low-level introductions and RDM-basics to elaborate and in-detail coding sessions. For three days we worked together, discussed use cases and new RDM tools. But we also enjoyed the social program such as a tour of the AiX Cave and the server room of the ITC’s High Performance Computing Center. There was a good balance between cognitively demanding workshop sessions and more relaxing social events and lunch breaks which hopefully resulted in an enjoyable and rewarding experience for all participants. On the very last day of the workshop we offered a session to especially discuss ideas and directions for the DKZ.2R. The feedback we got in this session will help us in moving forward with the DKZ.2R and making a lasting impact for future researchers.