Jens E. d’Hondt

Hi there! I’m Jens, a Postdoctoral Scientist at the Barcelona Supercomputing Center. In 2025, I defended my PhD thesis on “Efficient Algorithms for Multivariate Similarity Search” at TU Eindhoven, where I was part of the Data & AI cluster at the Department of Mathematics and Computer Science.

My research interests are twofold. Currently, I am working on super-resolution models for downscaling atmospheric simulations, with a focus on probabilistic models with physics-based constraints. In the past, I have worked on the design and implementation of efficient algorithms for multivariate similarity search, particularly in the context of high-dimensional data (i.e., vector search). This involved detection of high-order relationships in large databases, as well as the design of efficient algorithms for similarity search on multivariate data.

Some recent projects I have been working on include:

• Super-resolution models for downscaling atmospheric simulations using physics-based neural networks and ML-based simulation emulators, described in a preprint under review at Nature Communications and presented at EGU 2026.
• Fast detection of high-order relationships in large databases with the Correlation Detective library: VLDB 2022, VLDB Journal, and a call for multivariate similarity search algorithms at ICDE 2024. My PhD thesis brings this work together.
• Efficient subsequence search on multivariate time series: MS-Index (VLDB 2026) and MULISSE (ICDE 2025 workshop).
• Large-scale evaluations of distance measures for multivariate time series: a SIGMOD 2025 study of 30 measures, the earlier ICDE 2024 workshop paper, and a survey of over 100 time-series distance measures.
• Synthetic data that keeps the correlation structure of the original: Generative Correlation Manifolds.

All papers, with abstracts and links to code and slides, are on the publications page; recordings and slides are under talks. Longer, plain-language write-ups of this work are on the blog. A short CV is on the site, and a PDF version is available too.

Please feel free to reach out to me if you have any questions or would like to discuss any of my work. I am always open to new collaborations and opportunities.

News

• Sep 14, 2026: I started a blog. The first post is a practical guide to choosing a distance measure for multivariate time series, based on our SIGMOD 2025 study.
• Aug 7, 2026: New preprint alert! My paper “Neural downscaling of air-quality simulations requires structural correction before spatial refinement” is now available as a preprint and is currently under review at Nature Communications. Read it here.
• Mar 20, 2026: My paper on physics-based super-resolution models for downscaling atmospheric simulations has been accepted at EGU 2026 in Vienna, Austria. The abstract can be found here.
• Dec 16, 2025: I successfully defended my PhD thesis at TU Eindhoven!
• Dec 1, 2025: I have joined the Barcelona Supercomputing Center as a Postdoctoral Scientist in the Earth Sciences Department.
• Sep 16, 2025: Our paper on MS-Index, a novel algorithm for subsequence similarity search on multivariate time series, has been accepted at VLDB 2026 in Boston, USA. The paper and code can be found here.
• Sep 3, 2025: New paper alert! Download the preprint of my paper on synthetic data generation using generative correlation manifolds (GCM) here.
• June 23, 2025: Our paper on the evaluation of multivariate distance measures has been accepted at SIGMOD 2025 in Berlin, Germany. The paper is available here.
• May 19, 2025: I will be presenting our work on variable length multivariate time series search at the MulTiSa workshop at the International Conference on Data Engineering (ICDE) in Hong Kong.
• May 13, 2024: I will be presenting our work on evaluating multivariate distance measures at the MulTiSa workshop at the International Conference on Data Engineering (ICDE) in Utrecht, The Netherlands.
• Oct 11, 2023: Our paper on multivariate correlation analysis under different correlation measures has been accepted at the VLDB Journal. The paper can be found here.
• Mar 24, 2023: I will be giving a talk on multivariate correlation analysis at the Dutch Seminar on Data Systems Design (DSDSD) (online).
• Sep 20, 2022: I will be joining the consortium of the EU-funded project STELAR, where I will be working on the implementation of the Correlation Detective algorithm for analytics pipelines in the agri-food sector.
• Sep 9, 2022: My colleague Koen Minartz and I will be travelling to Sydney, Australia to present our work on Correlation Detective at VLDB 2022. The recording of our talk can be found here.
• Nov 17, 2019: I will be presenting our paper on automated health nudges at the International Conference on Modeling using Context (CONTEXT 2019) in Trento, Italy.
• Nov 12, 2019: I will be presenting our work on automated health nudges at the Data Science Summit 2019 in Eindhoven.
• April 25, 2019: TU/e’s news paper Cursor wrote an article about my work on automated health motivation systems and the randomized controlled trial we conducted at the university.