@article{21600,
  abstract     = {{Many problems in science and engineering require an efficient numerical approximation of integrals or solutions to differential equations. For systems with rapidly changing dynamics, an equidistant discretization is often inadvisable as it results in prohibitively large errors or computational effort. To this end, adaptive schemes, such as solvers based on Runge–Kutta pairs, have been developed which adapt the step size based on local error estimations at each step. While the classical schemes apply very generally and are highly efficient on regular systems, they can behave suboptimally when an inefficient step rejection mechanism is triggered by structurally complex systems such as chaotic systems. To overcome these issues, we propose a method to tailor numerical schemes to the problem class at hand. This is achieved by combining simple, classical quadrature rules or ODE solvers with data-driven time-stepping controllers. Compared with learning solution operators to ODEs directly, it generalizes better to unseen initial data as our approach employs classical numerical schemes as base methods. At the same time it can make use of identified structures of a problem class and, therefore, outperforms state-of-the-art adaptive schemes. Several examples demonstrate superior efficiency. Source code is available at https://github.com/lueckem/quadrature-ML.}},
  author       = {{Dellnitz, Michael and Hüllermeier, Eyke and Lücke, Marvin and Ober-Blöbaum, Sina and Offen, Christian and Peitz, Sebastian and Pfannschmidt, Karlson}},
  journal      = {{SIAM Journal on Scientific Computing}},
  number       = {{2}},
  pages        = {{A579--A595}},
  title        = {{{Efficient time stepping for numerical integration using reinforcement  learning}}},
  doi          = {{10.1137/21M1412682}},
  volume       = {{45}},
  year         = {{2023}},
}

@inproceedings{51209,
  author       = {{Hanselle, Jonas Manuel and Kornowicz, Jaroslaw and Heid, Stefan and Thommes, Kirsten and Hüllermeier, Eyke}},
  booktitle    = {{LWDA’23: Learning, Knowledge, Data, Analysis. }},
  editor       = {{Leyer, M and Wichmann, J}},
  issn         = {{1613-0073}},
  title        = {{{Comparing Humans and Algorithms in Feature Ranking: A Case-Study in the Medical Domain}}},
  year         = {{2023}},
}

@inproceedings{48778,
  author       = {{Muschalik, Maximilian and Fumagalli, Fabian and Jagtani, Rohit and Hammer, Barbara and Huellermeier, Eyke}},
  booktitle    = {{Proceedings of the World Conference on Explainable Artificial Intelligence (xAI)}},
  isbn         = {{9783031440632}},
  issn         = {{9783031440649}},
  title        = {{{iPDP: On Partial Dependence Plots in Dynamic Modeling Scenarios}}},
  doi          = {{10.1007/978-3-031-44064-9_11}},
  year         = {{2023}},
}

@inbook{48776,
  author       = {{Muschalik, Maximilian and Fumagalli, Fabian and Hammer, Barbara and Huellermeier, Eyke}},
  booktitle    = {{Machine Learning and Knowledge Discovery in Databases: Research Track - European Conference (ECML PKDD)}},
  isbn         = {{9783031434174}},
  issn         = {{1611-3349}},
  publisher    = {{Springer Nature Switzerland}},
  title        = {{{iSAGE: An Incremental Version of SAGE for Online Explanation on Data Streams}}},
  doi          = {{10.1007/978-3-031-43418-1_26}},
  year         = {{2023}},
}

@inproceedings{48775,
  author       = {{Fumagalli, Fabian and Muschalik, Maximilian and Hüllermeier, Eyke and Hammer, Barbara}},
  booktitle    = {{Proceedings of the European Symposium on Artificial Neural Networks (ESANN)}},
  location     = {{Bruges (Belgium) and online}},
  title        = {{{On Feature Removal for Explainability in Dynamic Environments}}},
  doi          = {{10.14428/ESANN/2023.ES2023-148}},
  year         = {{2023}},
}

@inproceedings{52230,
  author       = {{Fumagalli, Fabian and Muschalik, Maximilian and Kolpaczki, Patrick and Hüllermeier, Eyke and Hammer, Barbara}},
  booktitle    = {{Advances in Neural Information Processing Systems (NeurIPS)}},
  pages        = {{11515----11551}},
  title        = {{{SHAP-IQ: Unified Approximation of any-order Shapley Interactions}}},
  volume       = {{36}},
  year         = {{2023}},
}

@unpublished{30868,
  abstract     = {{Algorithm configuration (AC) is concerned with the automated search of the
most suitable parameter configuration of a parametrized algorithm. There is
currently a wide variety of AC problem variants and methods proposed in the
literature. Existing reviews do not take into account all derivatives of the AC
problem, nor do they offer a complete classification scheme. To this end, we
introduce taxonomies to describe the AC problem and features of configuration
methods, respectively. We review existing AC literature within the lens of our
taxonomies, outline relevant design choices of configuration approaches,
contrast methods and problem variants against each other, and describe the
state of AC in industry. Finally, our review provides researchers and
practitioners with a look at future research directions in the field of AC.}},
  author       = {{Schede, Elias and Brandt, Jasmin and Tornede, Alexander and Wever, Marcel Dominik and Bengs, Viktor and Hüllermeier, Eyke and Tierney, Kevin}},
  booktitle    = {{arXiv:2202.01651}},
  title        = {{{A Survey of Methods for Automated Algorithm Configuration}}},
  year         = {{2022}},
}

@inproceedings{32311,
  abstract     = {{Testing is one of the most frequent means of quality assurance for software. Property-based testing aims at generating test suites for checking code against user-defined properties. Test input generation is, however, most often independent of the property to be checked, and is instead based on random or user-defined data generation.In this paper, we present property-driven unit testing of functions with numerical inputs and outputs. Alike property-based testing, it allows users to define the properties to be tested for. Contrary to property-based testing, it also uses the property for a targeted generation of test inputs. Our approach is a form of learning-based testing where we first of all learn a model of a given black-box function using standard machine learning algorithms, and in a second step use model and property for test input generation. This allows us to test both predefined functions as well as machine learned regression models. Our experimental evaluation shows that our property-driven approach is more effective than standard property-based testing techniques.}},
  author       = {{Sharma, Arnab and Melnikov, Vitaly and Hüllermeier, Eyke and Wehrheim, Heike}},
  booktitle    = {{Proceedings of the 10th IEEE/ACM International Conference on Formal Methods in Software Engineering (FormaliSE)}},
  pages        = {{113--123}},
  publisher    = {{IEEE}},
  title        = {{{Property-Driven Testing of Black-Box Functions}}},
  year         = {{2022}},
}

@inproceedings{34542,
  author       = {{Campagner, Andrea and Lienen, Julian and Hüllermeier, Eyke and Ciucci, Davide}},
  booktitle    = {{Lecture Notes in Computer Science}},
  location     = {{Suzhou, China}},
  pages        = {{57--70}},
  publisher    = {{Springer}},
  title        = {{{Scikit-Weak: A Python Library for Weakly Supervised Machine Learning}}},
  volume       = {{13633}},
  year         = {{2022}},
}

@unpublished{31546,
  abstract     = {{In semi-supervised learning, the paradigm of self-training refers to the idea of learning from pseudo-labels suggested by the learner itself. Across various domains, corresponding methods have proven effective and achieve state-of-the-art performance. However, pseudo-labels typically stem from ad-hoc heuristics, relying on the quality of the predictions though without guaranteeing their validity. One such method, so-called credal self-supervised learning, maintains pseudo-supervision in the form of sets of (instead of single) probability distributions over labels, thereby allowing for a flexible yet uncertainty-aware labeling. Again, however, there is no justification beyond empirical effectiveness. To address this deficiency, we make use of conformal prediction, an approach that comes with guarantees on the validity of set-valued predictions. As a result, the construction of credal sets of labels is supported by a rigorous theoretical foundation, leading to better calibrated and less error-prone supervision for unlabeled data. Along with this, we present effective algorithms for learning from credal self-supervision. An empirical study demonstrates excellent calibration properties of the pseudo-supervision, as well as the competitiveness of our method on several benchmark datasets.}},
  author       = {{Lienen, Julian and Demir, Caglar and Hüllermeier, Eyke}},
  booktitle    = {{arXiv:2205.15239}},
  title        = {{{Conformal Credal Self-Supervised Learning}}},
  year         = {{2022}},
}

@unpublished{30867,
  abstract     = {{In online algorithm selection (OAS), instances of an algorithmic problem
class are presented to an agent one after another, and the agent has to quickly
select a presumably best algorithm from a fixed set of candidate algorithms.
For decision problems such as satisfiability (SAT), quality typically refers to
the algorithm's runtime. As the latter is known to exhibit a heavy-tail
distribution, an algorithm is normally stopped when exceeding a predefined
upper time limit. As a consequence, machine learning methods used to optimize
an algorithm selection strategy in a data-driven manner need to deal with
right-censored samples, a problem that has received little attention in the
literature so far. In this work, we revisit multi-armed bandit algorithms for
OAS and discuss their capability of dealing with the problem. Moreover, we
adapt them towards runtime-oriented losses, allowing for partially censored
data while keeping a space- and time-complexity independent of the time
horizon. In an extensive experimental evaluation on an adapted version of the
ASlib benchmark, we demonstrate that theoretically well-founded methods based
on Thompson sampling perform specifically strong and improve in comparison to
existing methods.}},
  author       = {{Tornede, Alexander and Bengs, Viktor and Hüllermeier, Eyke}},
  booktitle    = {{Proceedings of the 36th AAAI Conference on Artificial Intelligence}},
  publisher    = {{AAAI}},
  title        = {{{Machine Learning for Online Algorithm Selection under Censored Feedback}}},
  year         = {{2022}},
}

@unpublished{30865,
  abstract     = {{The problem of selecting an algorithm that appears most suitable for a
specific instance of an algorithmic problem class, such as the Boolean
satisfiability problem, is called instance-specific algorithm selection. Over
the past decade, the problem has received considerable attention, resulting in
a number of different methods for algorithm selection. Although most of these
methods are based on machine learning, surprisingly little work has been done
on meta learning, that is, on taking advantage of the complementarity of
existing algorithm selection methods in order to combine them into a single
superior algorithm selector. In this paper, we introduce the problem of meta
algorithm selection, which essentially asks for the best way to combine a given
set of algorithm selectors. We present a general methodological framework for
meta algorithm selection as well as several concrete learning methods as
instantiations of this framework, essentially combining ideas of meta learning
and ensemble learning. In an extensive experimental evaluation, we demonstrate
that ensembles of algorithm selectors can significantly outperform single
algorithm selectors and have the potential to form the new state of the art in
algorithm selection.}},
  author       = {{Tornede, Alexander and Gehring, Lukas and Tornede, Tanja and Wever, Marcel Dominik and Hüllermeier, Eyke}},
  booktitle    = {{Machine Learning}},
  title        = {{{Algorithm Selection on a Meta Level}}},
  year         = {{2022}},
}

@article{33090,
  abstract     = {{<jats:title>Abstract</jats:title><jats:p>Heated tool butt welding is a method often used for joining thermoplastics, especially when the components are made out of different materials. The quality of the connection between the components crucially depends on a suitable choice of the parameters of the welding process, such as heating time, temperature, and the precise way how the parts are then welded. Moreover, when different materials are to be joined, the parameter values need to be tailored to the specifics of the respective material. To this end, in this paper, three approaches to tailor the parameter values to optimize the quality of the connection are compared: a heuristic by Potente, statistical experimental design, and Bayesian optimization. With the suitability for practice in mind, a series of experiments are carried out with these approaches, and their capabilities of proposing well-performing parameter values are investigated. As a result, Bayesian optimization is found to yield peak performance, but the costs for optimization are substantial. In contrast, the Potente heuristic does not require any experimentation and recommends parameter values with competitive quality.</jats:p>}},
  author       = {{Gevers, Karina and Tornede, Alexander and Wever, Marcel Dominik and Schöppner, Volker and Hüllermeier, Eyke}},
  issn         = {{0043-2288}},
  journal      = {{Welding in the World}},
  keywords     = {{Metals and Alloys, Mechanical Engineering, Mechanics of Materials}},
  publisher    = {{Springer Science and Business Media LLC}},
  title        = {{{A comparison of heuristic, statistical, and machine learning methods for heated tool butt welding of two different materials}}},
  doi          = {{10.1007/s40194-022-01339-9}},
  year         = {{2022}},
}

@techreport{36227,
  author       = {{Hammer, Barbara and Hüllermeier, Eyke and Lohweg, Volker and Schneider, Alexander and Schenck, Wolfram and Kuhl, Ulrike and Braun, Marco and Pfeifer, Anton and Holst, Christoph-Alexander and Schmidt, Malte and Schomaker, Gunnar and Tornede, Tanja}},
  title        = {{{Schlussbericht ITS.ML: Intelligente Technische Systeme der nächsten Generation durch Maschinelles Lernen. Forschungsvorhaben zur automatisierten Analyse von Daten mittels Maschinellen Lernens}}},
  doi          = {{10.4119/unibi/2965622}},
  year         = {{2022}},
}

@article{48780,
  abstract     = {{Explainable Artificial Intelligence (XAI) has mainly focused on static learning tasks so far. In this paper, we consider XAI in the context of online learning in dynamic environments, such as learning from real-time data streams, where models are learned incrementally and continuously adapted over the course of time. More specifically, we motivate the problem of explaining model change, i.e. explaining the difference between models before and after adaptation, instead of the models themselves. In this regard, we provide the first efficient model-agnostic approach to dynamically detecting, quantifying, and explaining significant model changes. Our approach is based on an adaptation of the well-known Permutation Feature Importance (PFI) measure. It includes two hyperparameters that control the sensitivity and directly influence explanation frequency, so that a human user can adjust the method to individual requirements and application needs. We assess and validate our method’s efficacy on illustrative synthetic data streams with three popular model classes.}},
  author       = {{Muschalik, Maximilian and Fumagalli, Fabian and Hammer, Barbara and Huellermeier, Eyke}},
  issn         = {{0933-1875}},
  journal      = {{KI - Künstliche Intelligenz}},
  keywords     = {{Artificial Intelligence}},
  number       = {{3-4}},
  pages        = {{211--224}},
  publisher    = {{Springer Science and Business Media LLC}},
  title        = {{{Agnostic Explanation of Model Change based on Feature Importance}}},
  doi          = {{10.1007/s13218-022-00766-6}},
  volume       = {{36}},
  year         = {{2022}},
}

@article{24143,
  author       = {{Drees, Jan Peter and Gupta, Pritha and Hüllermeier, Eyke and Jager, Tibor and Konze, Alexander and Priesterjahn, Claudia and Ramaswamy, Arunselvan and Somorovsky, Juraj}},
  journal      = {{14th ACM Workshop on Artificial Intelligence and Security}},
  title        = {{{Automated Detection of Side Channels in Cryptographic Protocols: DROWN the ROBOTs!}}},
  year         = {{2021}},
}

@article{24148,
  author       = {{Ramaswamy, Arunselvan and Hüllermeier, Eyke}},
  journal      = {{IEEE Transactions on Artificial Intelligence (to appear)}},
  title        = {{{Deep Q-Learning: Theoretical Insights from an Asymptotic Analysis}}},
  year         = {{2021}},
}

@article{21004,
  abstract     = {{Automated machine learning (AutoML) supports the algorithmic construction and data-specific customization of machine learning pipelines, including the selection, combination, and parametrization of machine learning algorithms as main constituents. Generally speaking, AutoML approaches comprise two major components: a search space model and an optimizer for traversing the space. Recent approaches have shown impressive results in the realm of supervised learning, most notably (single-label) classification (SLC). Moreover, first attempts at extending these approaches towards multi-label classification (MLC) have been made. While the space of candidate pipelines is already huge in SLC, the complexity of the search space is raised to an even higher power in MLC. One may wonder, therefore, whether and to what extent optimizers established for SLC can scale to this increased complexity, and how they compare to each other. This paper makes the following contributions: First, we survey existing approaches to AutoML for MLC. Second, we augment these approaches with optimizers not previously tried for MLC. Third, we propose a benchmarking framework that supports a fair and systematic comparison. Fourth, we conduct an extensive experimental study, evaluating the methods on a suite of MLC problems. We find a grammar-based best-first search to compare favorably to other optimizers.}},
  author       = {{Wever, Marcel Dominik and Tornede, Alexander and Mohr, Felix and Hüllermeier, Eyke}},
  issn         = {{0162-8828}},
  journal      = {{IEEE Transactions on Pattern Analysis and Machine Intelligence}},
  keywords     = {{Automated Machine Learning, Multi Label Classification, Hierarchical Planning, Bayesian Optimization}},
  pages        = {{1--1}},
  title        = {{{AutoML for Multi-Label Classification: Overview and Empirical Evaluation}}},
  doi          = {{10.1109/tpami.2021.3051276}},
  year         = {{2021}},
}

@article{21092,
  abstract     = {{Automated Machine Learning (AutoML) seeks to automatically find so-called machine learning pipelines that maximize the prediction performance when being used to train a model on a given dataset. One of the main and yet open challenges in AutoML is an effective use of computational resources: An AutoML process involves the evaluation of many candidate pipelines, which   are costly but often ineffective because they are canceled due to a timeout.
In this paper, we present an approach to predict the runtime of two-step machine learning pipelines with up to one pre-processor, which can be used to anticipate whether or not a pipeline will time out. Separate runtime models are trained offline for each algorithm that may be used in a pipeline, and an overall prediction is derived from these models. We empirically show that the approach increases successful evaluations made by an AutoML tool while preserving or even improving on the previously best solutions.}},
  author       = {{Mohr, Felix and Wever, Marcel Dominik and Tornede, Alexander and Hüllermeier, Eyke}},
  journal      = {{IEEE Transactions on Pattern Analysis and Machine Intelligence}},
  publisher    = {{IEEE}},
  title        = {{{Predicting Machine Learning Pipeline Runtimes in the Context of Automated Machine Learning}}},
  year         = {{2021}},
}

@inproceedings{21570,
  author       = {{Tornede, Tanja and Tornede, Alexander and Wever, Marcel Dominik and Hüllermeier, Eyke}},
  booktitle    = {{Proceedings of the Genetic and Evolutionary Computation Conference}},
  title        = {{{Coevolution of Remaining Useful Lifetime Estimation Pipelines for Automated Predictive Maintenance}}},
  year         = {{2021}},
}