Institute of Data Analysis and Process Design (IDP)
We create value from data
We use advanced data-based methods to create innovative solutions for business and industry. We address real-life challenges with scientific methods and a strong commitment to practicability. We are the leading educator and partner of choice for applied data science and business engineering in Switzerland.
Research Groups

Advanced scientific tools for solutions in the financial industry

Health and Envrionmental Analytics
Health and Environmental Analytics
Analyzing data to derive interpretable results using statistical and machine learning techniques

Maintenance, Mobility, AI & Society
Leverage AI and advanced modeling for innovations in predictive maintenance, mobility solutions, and socially aligned systems

Generating insights, creating value and fostering innovation in business processes and services

Visual Intelligence and Applications
As visual data becomes one of the most abundant and complex sources of information, Visual Intelligence is a key pillar of modern data science — enabling new ways to analyze, model, and communicate through images, video, and immersive environments
For Students
Consulting Services
News
Team IDP
Publikationen
-
Lüscher, Mila Francesca; Zgraggen, Jannik; Guo, Yuyan; Notaristefano, Antonio; Goren Huber, Lilach,
2024.
In:
Do, Phuc; Ezhilarasu, Cordelia, eds.,
Proceedings of the PHM Society European Conference 2024.
8th European Conference of the Prognostics and Health Management Society (PHME), Prague, Czech Republic, 3-5 July 2024.
PHM Society.
pp. 286-293.
Available from: https://doi.org/10.36001/phme.2024.v8i1.4059
-
Baumann, Joachim; Sapiezynski, Piotr; Heitz, Christoph; Hannák, Anikó,
2024.
Fairness in online ad delivery[paper].
In:
Proceedings of the 2024 ACM Conference on Fairness, Accountability, and Transparency.
7th ACM Conference on Fairness, Accountability, and Transparency (FAccT), Rio de Janeiro, Brazil, 3-6 June 2024.
Association for Computing Machinery.
pp. 1418-1432.
Available from: https://doi.org/10.1145/3630106.3658980
-
2024.
Journal of Scheduling.
27(3), pp. 257-275.
Available from: https://doi.org/10.1007/s10951-023-00787-5
-
Wulf, Jochen; Meierhofer, Jürg,
2024.
Utilizing large language models for automating technical customer support.
arXiv.
Available from: https://doi.org/10.48550/arXiv.2406.01407
-
2024.
A data-centric-AI trick to clean your dirty data.
In:
Detecting unusual or abnormal patterns in data is one of the common tasks of AI algorithms in commercial applications. In some applications, such as fraud detection, defect detection or medical diagnostics, anomaly detection is the main objective. In other applications, detecting abnormal data points is part of the data cleaning and preparation pipeline. In all cases, the use of AI-based methods relies on having a training dataset which can represent the normal behaviour, and must therefore be free of anomalies. Problems arise when we realize that having an anomaly-free training dataset is not always possible in practice: most real-world datasets are contaminated with unknown anomalies or mislabeled data..