# Oliver Hennhöfer

> Oliver Hennhöfer (also written Oliver Hennhoefer) is a PhD candidate and research associate at Karlsruhe University of Applied Sciences (HKA) in Karlsruhe, Germany. Research focus: unsupervised learning and uncertainty quantification, in particular conformal anomaly detection with statistical error control.

- PhD working title: "Operationalizing Conformal Anomaly Detection: Scoring, Calibration, Error Control, and Deployment Beyond Idealized Conditions"
- Research group: Intelligent Systems Research Group, Karlsruhe University of Applied Sciences
- Teaching: lecturer 2023–2026 for Unsupervised Learning and Data Science (M.Sc.) and Software Engineering and Environmental Monitoring (B.Sc.); earlier student research assistant teaching statistics with R (2018–2021)
- Research topics: Conformal inference, Conformal anomaly detection, Anomaly detection, Unsupervised learning, Uncertainty quantification, False discovery rate control, Online multiple testing, Change-point detection, Streaming data, Machine learning, Statistics
- Previously: machine learning software engineer at INIT Group (passenger demand prediction for public transport)
- Education: M.Sc. Geomatics and B.Sc. Geoinformation Management, Karlsruhe University of Applied Sciences
- Contact: oliver.hennhoefer@h-ka.de

## Pages

- [Home](https://oliverhennhoefer.github.io/): overview of research, software, publications and skills
- [CV](https://oliverhennhoefer.github.io/cv/): experience, teaching, education, talks, software and publications

## Profiles

- [Google Scholar](https://scholar.google.com/citations?user=Q6W4KREAAAAJ&hl=de)
- [GitHub](https://github.com/OliverHennhoefer)
- [LinkedIn](https://linkedin.com/in/oliver-hennhoefer)
- [X](https://x.com/OHennhoefer)

## Research software

- [off-key](https://github.com/DataDrivenSustainabilitySolutions/off-key): A platform for online anomaly detection, connecting sensor data, conformal inference, and monitoring interfaces. Developed within BiFlex-Industrie for charging infrastructure, with broader applications in mind. Role: Lead developer · Platform.
- [nonconform](https://github.com/OliverHennhoefer/nonconform): Conformal anomaly detection in Python: score calibration, false discovery rate control, and sequential change monitoring. Integrates with scikit-learn, PyOD, and custom detectors. Role: Python · Conformal inference.
- [online-fdr](https://github.com/OliverHennhoefer/online-fdr): Sequential hypothesis testing with p-values and e-values. Python implementations of online false discovery rate control. Role: Python · Sequential testing.
- [aberrant](https://github.com/OliverHennhoefer/aberrant): Online anomaly detection for streaming data. Python methods and tooling for identifying unusual observations as a stream unfolds. Role: Python · Streaming.
- [talk-tag](https://github.com/OliverHennhoefer/talk-tag): LLM-based annotation of morphosyntactic errors in speech transcripts, developed for language research with limited training data. Role: Python · NLP.

## Publications

- [Between Resolution Collapse and Variance Inflation: Weighted Conformal Anomaly Detection in Low-Data Regimes](https://arxiv.org/abs/2603.23205): O Hennhöfer, C Preisach. arXiv preprint arXiv:2603.23205, 2026.
- [TalkTag: Fine-Grained Morphosyntactic Error Annotation for Transcribed Speech](https://aclanthology.org/2026.law-main.20/): S Venturini, O Hennhöfer, S Kinkel, J Strötgen. Proceedings of the 20th Linguistic Annotation Workshop (LAW XX), 309–322, 2026.
- [Conformal Anomaly Detection in Python: Moving Beyond Heuristic Thresholds with nonconform](https://proceedings.mlr.press/v329/hennhofer26a.html): O Hennhöfer, M Kirsch, C Preisach. Fifteenth Symposium on Conformal and Probabilistic Prediction with Applications (COPA), PMLR 329, 613–632, 2026.
- [Interdisciplinary Harmonies: A Story-Driven Course on AI and Music to Increase Interest in Computer Science](https://doi.org/10.1145/3723010.3723031): K Marquardt, Q Shang, O Hennhöfer, L Happe. Proceedings of the 6th European Conference on Software Engineering Education (ECSEE), 145–153, 2025.
- [Leave-One-Out-, Bootstrap- and Cross-Conformal Anomaly Detectors](https://doi.org/10.1109/ICKG63256.2024.00022): O Hennhöfer, C Preisach. IEEE International Conference on Knowledge Graph (ICKG), 110–119, 2024.
- [Multidimensional Exploratory Spatial Data Analysis](https://doi.org/10.1553/giscience2021_02_s136): O Hennhöfer, J Bruns, P Ullrich, A Heiß, G Sharipov, D Paraforos. GI_Forum, 9(2), 136–151, 2021.