no code implementations • 14 Feb 2024 • Mira Jürgens, Nis Meinert, Viktor Bengs, Eyke Hüllermeier, Willem Waegeman
Trustworthy ML systems should not only return accurate predictions, but also a reliable representation of their uncertainty.
no code implementations • 22 Dec 2023 • Timo Kaufmann, Paul Weng, Viktor Bengs, Eyke Hüllermeier
Reinforcement learning from human feedback (RLHF) is a variant of reinforcement learning (RL) that learns from human feedback instead of relying on an engineered reward function.
no code implementations • 2 Dec 2023 • Yusuf Sale, Viktor Bengs, Michele Caprio, Eyke Hüllermeier
In the past couple of years, various approaches to representing and quantifying different types of predictive uncertainty in machine learning, notably in the setting of classification, have been proposed on the basis of second-order probability distributions, i. e., predictions in the form of distributions on probability distributions.
no code implementations • 1 Oct 2023 • Viktor Bengs, Björn Haddenhorst, Eyke Hüllermeier
We consider the task of identifying the Copeland winner(s) in a dueling bandits problem with ternary feedback.
no code implementations • 1 Feb 2023 • Jasmin Brandt, Marcel Wever, Dimitrios Iliadis, Viktor Bengs, Eyke Hüllermeier
Hyperparameter optimization (HPO) is concerned with the automated search for the most appropriate hyperparameter configuration (HPC) of a parameterized machine learning algorithm.
no code implementations • 1 Feb 2023 • Patrick Kolpaczki, Viktor Bengs, Maximilian Muschalik, Eyke Hüllermeier
The Shapley value, which is arguably the most popular approach for assigning a meaningful contribution value to players in a cooperative game, has recently been used intensively in explainable artificial intelligence.
no code implementations • 30 Jan 2023 • Viktor Bengs, Eyke Hüllermeier, Willem Waegeman
In this paper, we generalise these findings and prove a more fundamental result: There seems to be no loss function that provides an incentive for a second-order learner to faithfully represent its epistemic uncertainty in the same manner as proper scoring rules do for standard (first-order) learners.
1 code implementation • 1 Dec 2022 • Jasmin Brandt, Elias Schede, Viktor Bengs, Björn Haddenhorst, Eyke Hüllermeier, Kevin Tierney
We study the algorithm configuration (AC) problem, in which one seeks to find an optimal parameter configuration of a given target algorithm in an automated way.
no code implementations • 20 May 2022 • Thomas Mortier, Viktor Bengs, Eyke Hüllermeier, Stijn Luca, Willem Waegeman
In this paper, we extend the notion of calibration, which is commonly used to evaluate the validity of the aleatoric uncertainty representation of a single probabilistic classifier, to assess the validity of an epistemic uncertainty representation obtained by sets of probabilistic classifiers.
no code implementations • 11 Mar 2022 • Viktor Bengs, Eyke Hüllermeier, Willem Waegeman
Uncertainty quantification has received increasing attention in machine learning in the recent past.
no code implementations • 9 Feb 2022 • Viktor Bengs, Aadirupa Saha, Eyke Hüllermeier
In every round of the sequential decision problem, the learner makes a context-dependent selection of two choice alternatives (arms) to be compared with each other and receives feedback in the form of noisy preference information.
no code implementations • 9 Feb 2022 • Jasmin Brandt, Viktor Bengs, Björn Haddenhorst, Eyke Hüllermeier
We consider the combinatorial bandits problem with semi-bandit feedback under finite sampling budget constraints, in which the learner can carry out its action only for a limited number of times specified by an overall budget.
no code implementations • 3 Feb 2022 • Elias Schede, Jasmin Brandt, Alexander Tornede, Marcel Wever, Viktor Bengs, Eyke Hüllermeier, Kevin Tierney
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.
no code implementations • 2 Feb 2022 • Patrick Kolpaczki, Viktor Bengs, Eyke Hüllermeier
We propose the $\mathrm{Beat\, the\, Winner\, Reset}$ algorithm and prove a bound on its expected binary weak regret in the stationary case, which tightens the bound of current state-of-art algorithms.
1 code implementation • NeurIPS 2021 • Björn Haddenhorst, Viktor Bengs, Eyke Hüllermeier
The reliable identification of the “best” arm while keeping the sample complexity as low as possible is a common task in the field of multi-armed bandits.
1 code implementation • 13 Sep 2021 • Alexander Tornede, Viktor Bengs, Eyke Hüllermeier
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.
no code implementations • 2 Nov 2020 • Viktor Bengs, Eyke Hüllermeier
We consider a resource-aware variant of the classical multi-armed bandit problem: In each round, the learner selects an arm and determines a resource limit.
no code implementations • 11 Feb 2020 • Adil El Mesaoudi-Paul, Viktor Bengs, Eyke Hüllermeier
We consider an extension of the contextual multi-armed bandit problem, in which, instead of selecting a single alternative (arm), a learner is supposed to make a preselection in the form of a subset of alternatives.
no code implementations • ICML 2020 • Viktor Bengs, Eyke Hüllermeier
To formalize this goal, we introduce a reasonable notion of regret and derive lower bounds on the expected regret.
no code implementations • 30 Jul 2018 • Viktor Bengs, Robert Busa-Fekete, Adil El Mesaoudi-Paul, Eyke Hüllermeier
The aim of this paper is to provide a survey of the state of the art in this field, referred to as preference-based multi-armed bandits or dueling bandits.