InstitutTeam
Alexander Tornede

Alexander Tornede, M. Sc.

Alexander Tornede, M. Sc.
Adresse
Appelstraße 9a
30167 Hannover
Gebäude
Raum
Alexander Tornede, M. Sc.
Adresse
Appelstraße 9a
30167 Hannover
Gebäude
Raum

My Vision

Artificial intelligence is no longer science fiction, but can be found all around us. Be it the personal assistant on your smartphone, the driving assistant of your car, or the setup of your favorite clothing store - chances are that all of these leverage AI and/or have been created with the help of AI. As such, it might seem that AI is democratized and more or less easily available as a supporting tool for researchers and businesses. However, the majority of AI solutions present in our daily life has been developed by and are only available to large companies due to the associated personnel costs of employing data scientists.

Considering the impressive advances AI achieved in the recent years in fields such as natural language processing, image processing and product recommendation, this is unfortunate. Instead, I believe that everyone should be enabled to leverage these approaches to improve the world we live in. Therefore, my goal is to democratize AI and in particular machine learning (ML) by research in the area of automated machine learning (AutoML). I aim at enabling researchers from other disciplines and companies, which might not be able to afford large data science teams, to leverage ML through the means of AutoML tools.

To achieve this, I research on interactive and explainable AutoML targeted at supporting data scientists and domain experts in leveraging ML for their particular application in an intuitive, human-centered and explainable way. Together with my awesome colleagues, we work hard to make existing AutoML tools easier to use and understand by non-experts and in particular improve the understandability of the underlying AutoML process. We hope that this will demystify AutoML, such that non-expert users can interact with AutoML tools by injecting domain knowledge and tailoring the search process to their needs.

Research Interests

My interests are centered around different topics in machine learning and artificial intelligence, and in particular the automation thereof, called automated machine learning (AutoML). I am most interested in the following topics:

  • Interactive and Explainable AutoML
  • Green AutoML
  • Algorithm Selection
  • Applications to health/medicine

Curriculum Vitae

Publications

  • Marcel Wever, Alexander Tornede, Felix Mohr, Eyke Hüllermeier (2021): AutoML for Multi-Label Classification: Overview and Empirical EvaluationIEEE Transactions on Pattern Analysis and Machine Intelligence (TPAMI)
    DOI: 10.1109/tpami.2021.3051276
  • Alexander Tornede, Marcel Wever, Eyke Hüllermeier (2020): Extreme Algorithm Selection with Dyadic Feature RepresentationDiscovery Science 2020
    DOI: 10.1007/978-3-030-61527-7_21
    arXiv: 2001.10741
  • Felix Mohr, Marcel Wever, Alexander Tornede, Eyke Hüllermeier (2021): Predicting Machine Learning Pipeline Runtimes in the Context of Automated Machine LearningIEEE Transactions on Pattern Analysis and Machine Intelligence (TPAMI)
    DOI: 10.1109/TPAMI.2021.3056950
  • Alexander Tornede, Marcel Wever, Stefan Werner, Felix Mohr, Eyke Hüllermeier (2020): Run2Survive: A Decision-theoretic Approach to Algorithm Selection based on Survival Analysis12th Asian Conference on Machine Learning (ACML)
    DOI: 10.48550/arXiv.2007.02816
    arXiv: 2007.02816
  • Alexander Tornede, Marcel Wever, Eyke Hüllermeier (2019): Algorithm Selection as Recommendation: From Collaborative Filtering to Dyad Ranking29th Workshop Computational Intelligence
  • Jonas Hanselle, Alexander Tornede, Marcel Wever, Eyke Hüllermeier (2020): Hybrid Ranking and Regression for Algorithm Selection43rd German Conference on Artificial Intelligence (KI 2020)
    DOI: 10.1007/978-3-030-58285-2_5
  • Tanja Tornede, Alexander Tornede, Marcel Wever, Felix Mohr, Eyke Hüllermeier (2021): AutoML for Predictive Maintenance: One Tool to RUL them allIoTStream @ ECMLPKDD 2020, 2020
    DOI: 10.1007/978-3-030-66770-2_8
  • Marcel Wever, Felix Mohr, Alexander Tornede, Eyke Hüllermeier (2019): Automating Multi-Label Classification Extending ML-Plan6th ICML Workshop on Automated Machine Learning
  • Tanja Tornede, Alexander Tornede, Marcel Wever, Eyke Hüllermeier (2021): Coevolution of Remaining Useful Lifetime Estimation Pipelines for Automated Predictive MaintenanceProceedings of the Genetic and Evolutionary Computation Conference (GECCO 2021)
    DOI: 10.1145/3449639.3459395
  • Alexander Tornede, Lukas Gehring, Tanja Tornede, Marcel Wever, Eyke Hüllermeier (2022): Algorithm selection on a meta levelMachine Learning
    DOI: 10.1007/s10994-022-06161-4
  • Tanja Tornede, Alexander Tornede, Jonas Hanselle, Marcel Wever, Felix Mohr, Eyke Hüllermeier (2021): Towards Green Automated Machine Learning: Status Quo and Future DirectionsArXiv
    arXiv: 2111.05850
  • Alexander Tornede, Marcel Wever, Eyke Hüllermeier (2020): Towards Meta-Algorithm SelectionWorkshop on Meta-Learning (MetaLearn 2020) @ NeurIPS 2020
    arXiv: 2011.08784
  • Marcel Wever, Alexander Tornede, Felix Mohr, Eyke Hüllermeier (2020): LiBRe: Label-Wise Selection of Base Learners in Binary Relevance for Multi-Label ClassificationSymposium on Intelligent Data Analysis (IDA 2020)
    DOI: 10.1007/978-3-030-44584-3_44
  • Elias Schede, Jasmin Brandt, Alexander Tornede, Marcel Wever, Viktor Bengs, Eyke Hüllermeier, Kevin Tierney (2022): A Survey of Methods for Automated Algorithm ConfigurationJournal of Artificial Intelligence Research
    DOI: 10.1613/jair.1.13676
  • Eyke Hüllermeier, Felix Mohr, Alexander Tornede, Marcel Wever (2021): Automated Machine Learning, Bounded Rationality, and Rational MetareasoningECMLPKDD Workshop on Automating Data Science (ADS2021)
    DOI: 2109.04744
  • Felix Mohr, Marcel Wever, Alexander Tornede, Eyke Hüllermeier (2019): From Automated to On-The-Fly Machine LearningINFORMATIK 2019: 50 Jahre Gesellschaft für Informatik–Informatik für Gesellschaft
    DOI: 10.18420/inf2019_40
  • Karina Gevers, Alexander Tornede, Marcel Wever, Volker Schöppner, Eyke Hüllermeier (2022): A comparison of heuristic, statistical, and machine learning methods for heated tool butt welding of two different materialsWelding in the World
    DOI: 10.1007/s40194-022-01339-9
  • Lukas Fehring, Jonas Hanselle, Alexander Tornede (2022): HARRIS: Hybrid Ranking and Regression Forests for Algorithm SelectionWorkshop on Meta-Learning (MetaLearn 2022) @ NeurIPS 2022
    arXiv: 2210.17341
  • Alexander Tornede, Viktor Bengs, Eyke Hüllermeier (2022): Machine Learning for Online Algorithm Selection under Censored FeedbackProceedings of the 36th AAAI Conference on Artificial Intelligence (AAAI 2022)
    DOI: 10.1609/aaai.v36i9.21279
  • Jonas Hanselle, Alexander Tornede, Marcel Wever, Eyke Hüllermeier (2021): Algorithm Selection as Superset Learning: Constructing Algorithm Selectors from Imprecise Performance Data25th Pacific-Asia Conference on Knowledge Discovery and Data Mining (PAKDD-2021)
    DOI: 10.1007/978-3-030-75762-5_13

Projects

  • ERC Starting Grant: Interactive and Explainable Human-Centered AutoML
    Trust and interactivity are key factors in the future development and use of automated machine learning (AutoML), supporting developers and researchers in determining powerful task-specific machine learning pipelines, including pre-processing, predictive algorithm, their hyperparameters and--if applicable--the architecture design of deep neural networks. Although AutoML is ready for its prime time after it achieved impressive results in several machine learning (ML) applications and its efficiency improved by several orders of magnitudes in recent years, democratization of machine learning via AutoML is still not achieved. In contrast to previously purely automation-centered approaches, ixAutoML is designed with human users at its heart in several stages. First of all, the foundation of trustful use of AutoML will be based on explanations of its results and processes. Therefore, we aim for: (i) Explaining static effects of design decisions in ML pipelines optimized by state-of-the-art AutoML systems. (ii) Explaining dynamic AutoML policies for temporal aspects of dynamically adapted hyperparameters while ML models are trained. These explanations will be the base for allowing interactions, bringing the best of two worlds together: human intuition and generalization capabilities for complex systems, and efficiency of systematic optimization approaches for AutoML. Concretely, we aim for: (iii) Enabling interactions between humans and AutoML by taking human's latent knowledge into account and learning when to interact. (iv) Building first ixAutoML prototypes and showing its efficiency in the context of Industry 4.0. Perfectly aligned with the EU's AI strategy and recent efforts on interpretability in the ML community, we strongly believe that this timely human-centered ixAutoML will have a substantial impact on the democratization of machine learning.
    Leitung: Prof. Dr. Marius Lindauer
    Team: AutoML
    Jahr: 2022
    Förderung: EU
    Laufzeit: 2022-2027