Search our Publications
Latest Publications
Christiane Attig, Patricia Wollstadt, Thomas Franke, Tim Schrills, Christiane Wiebel , "More than Task Performance: Developing New Criteria for Successful Human-AI Teaming Using the Cooperative Card Game Hanabi", CHI EA '24: Extended Abstracts of the CHI Conference on Human Factors in Computing Systems, no. 245, pp. 1-11, 2024.
AbstractAs we shift to designing AI agents as teammates rather than tools, the social aspects of human-AI interaction become more pronounced. Consequently, to develop agents that are able to navigate the social dynamics that accompany cooperative teamwork, evaluation criteria that refer only to objective task performance will not be suffi cient. We propose perceived cooperativity and teaming perception as subjective metrics for investigating successfu...
Download Bibtex file Download PDF
Christian Internó, Barbara Hammer, Yaochu Jin, Markus Olhofer , "FedLEx: Federated Loss Exploration for Improved Convergence", Machine Learning Summer School in Okinawa 2024, 2024.
AbstractFederated Learning (FL) offers a decentralized machine learning framework, allowing participants to collaboratively train models while keeping data localized. In non-IID settings, where data distribution among clients isn’t consistent, challenges arise that hinder global model convergence and good generalization. To alleviate this, we introduce the Federated Loss Exploration (FedLEx) method. FedLEx incorporates a loss landscape exploration phase ...
Download Bibtex file Per Mail Request
Johannes Varga, Guenther Raidl, Tobias Rodemann , "Selecting User Queries in Interactive Job Scheduling", Eurocast 2024, 19th International Conference on Computer Aided Systems Theory, 2024.
AbstractWe consider a class of job scheduling problems in which human users, e.g., the personnel of a company, need to perform jobs on some shared machines and the availabilities of these users as well as the machines is critical. In such situations it is rarely practical to ask users to fully specify their availability times. Instead we assume users initially only propose a single starting time for each of their jobs, and a feasible and optimized schedu...
Download Bibtex file Per Mail Request
Christian Internó, Markus Olhofer, Yaochu Jin, Barbara Hammer , "Federated Loss Exploration for Improved Convergence on non-IID data", International Joint Conference on Neural Networks (IJCNN), 2024.
AbstractFederated learning (FL) has emerged as a groundbreaking paradigm in machine learning (ML), offering privacy-preserving collaborative model training across diverse datasets. Despite its promise, FL faces significant hurdles in non-identically and independently distributed (non-IID) data scenarios, where most existing methods often struggle with data heterogeneity and lack robustness in performance. This paper introduces Federated Loss Exploration ...
Download Bibtex file Download PDF
Christian Internó, Elena Raponi, Niki van Stein, Thomas Bäck, Markus Olhofer, Yaochu Jin, Barbara Hammer , "Automated Federated Learning via Informed Pruning", International Conference on Automated Machine Learning (AUTOML 24), 2024.
AbstractFederated learning (FL) represents a pivotal shift in machine learning (ML) as it enables collaborative training of local ML models coordinated by a central aggregator, all without the need to exchange local data. However, its application on edge devices is hindered by limited computational capabilities and data communication challenges, compounded by the inherent complexity of Deep Learning (DL) models. Model pruning is identified as a key tec...
Download Bibtex file Download PDF
Lydia Fischer and Patricia Wollstadt , "Precision and Recall Reject Curves for Classification", WSOM, 2024.
AbstractFor some classification scenarios, it is desirable to use only those classification instances that a trained model associates with a high certainty. To obtain such high-certainty instances, previous work has proposed accuracy-reject curves. Reject curves allow to evaluate and compare the performance of different certainty measures over a range of thresholds for accepting or rejecting classifications. However, the accuracy may not be the most sui...
Download Bibtex file Per Mail Request
Duc Anh Nguyen , "Efficient tuning of automated machine learning pipelines", Leiden University, 2024.
AbstractAutoML has attracted community attention due to its success in shortening the machine learning development cycle for real-world applications. Optimization plays a crucial role in AutoML frameworks by helping to identify a fine-tuned ML pipeline that suits a given practical problem. Several state-of-the-art optimization approaches, including Bayesian optimization, Bandit learning, and Racing procedures, have been proposed to enhance the performanc...
Download Bibtex file Per Mail Request
Johannes Varga, Guenther Raidl, Elina Rönnberg, Tobias Rodemann , "Scheduling jobs using queries to interactively learn human availability times", Computers & Operations Research, vol. 167, pp. 106648, 2024.
AbstractThe solution to a job scheduling problem that involves humans as well some other shared resource has to consider the humans’ availability times. For practical acceptance of a scheduling tool, it is crucial that the interaction with the humans is kept simple and to a minimum. It is rarely practical to ask users to fully specify their availability times or to let them enumerate all possible starting times for their jobs. In the scenario we are con...
Download Bibtex file Download PDF
Daniel Tanneberg, Felix Ocker, Stephan Hasler, Jörg Deigmöller, Anna Belardinelli, Chao Wang, Heiko Wersing, Bernhard Sendhoff, Michael Gienger , "To Help or Not to Help: LLM-based Attentive Support for Human-Robot Group Interactions", Late Breaking Work Poster - IEEE International Conference on Robotics and Automation (ICRA), 2024.
AbstractHumans are inherently social beings. To seamlessly integrate robots into our daily lives, it is crucial that they can engage in multiparty interactions effectively and supportively, without disrupting group dynamics. Hence, we asked “How can a robot provide unobtrusive physical support within a group of humans?”...
Download Bibtex file Per Mail Request
Tim Puphal, Ryohei Hirano, Akihito Kimata, Julian Eggert , "Reducing Warning Errors in Driver Support with Personalized Risk Maps", IEEE International Conference on Vehicular Electronics and Safety 2024, 2024.
AbstractWe consider the problem of human-focused driver support. State-of-the-art personalization concepts allow to estimate parameters for vehicle control systems or driver models. However, there are currently few approaches proposed that use personalized models and evaluate the effectiveness in the form of general risk warning. In this paper, we therefore propose a warning system that estimates a personalized risk factor for the given driver based on t...
Download Bibtex file Per Mail Request
1 ... 7 8 9 10 11 12 ... 170
Follow us
