Research groups / ISMLL

ISMLL

Information Systems & Machine Learning Lab

Machine learning that ships — recommender systems, time series, AutoML, deep learning, neuro-symbolic and high-dimensional ML.

Lead: Prof. Dr. Dr. Lars Schmidt-Thieme · Prof. Dr. Tom Hanika Group website ↗

ISMLL is an internationally recognised machine-learning lab. It is the birthplace of Bayesian Personalised Ranking (BPR) and Factorization Machines, and holds the 2025 Web Conference Test-of-Time Award.

Research

  • Recommender systems & personalisation
  • Time-series forecasting, including irregularly-sampled and foundation-model approaches
  • AutoML, hyperparameter optimisation and meta-learning
  • Deep learning for tabular and graph-structured data; model compression for LLMs
  • Neuro-symbolic and high-dimensional machine learning

ISMLL runs the English-taught, tuition-free International M.Sc. in Data Analytics.