@article{63498, author = {{Kirchgässner, Wilhelm and Förster, Nikolas and Piepenbrock, Till and Schweins, Oliver and Wallscheid, Oliver}}, journal = {{IEEE Transactions on Power Electronics}}, keywords = {{Mathematical models, Estimation, Data models, Convolutional neural networks, Accuracy, Magnetic hysteresis, Magnetic cores, Temperature measurement, Magnetic domains, Temperature distribution, Convolutional neural network (CNN), machine learning (ML), magnetics}}, number = {{2}}, pages = {{3326--3335}}, title = {{{HARDCORE: H-Field and Power Loss Estimation for Arbitrary Waveforms With Residual, Dilated Convolutional Neural Networks in Ferrite Cores}}}, doi = {{10.1109/TPEL.2024.3488174}}, volume = {{40}}, year = {{2025}}, } @inproceedings{62078, abstract = {{Fiber reinforced plastics (FRP) exhibit strongly non-linear deformation behavior. To capture this in simulations, intricate models with a variety of parameters are typically used. The identification of values for such parameters is highly challenging and requires in depth understanding of the model itself. Machine learning (ML) is a promising approach for alleviating this challenge by directly predicting parameters based on experimental results. So far, this works mostly for purely artificial data. In this work, two approaches to generalize to experimental data are investigated: a sequential approach, leveraging understanding of the constitutive model and a direct, purely data driven approach. This is exemplary carried out for a highly non-linear strain rate dependent constitutive model for the shear behavior of FRP.The sequential model is found to work better on both artificial and experimental data. It is capable of extracting well suited parameters from the artificial data under realistic conditions. For the experimental data, the model performance depends on the composition of the experimental curves, varying between excellently suiting and reasonable predictions. Taking the expert knowledge into account for ML-model training led to far better results than the purely data driven approach. Robustifying the model predictions on experimental data promises further improvement. }}, author = {{Gerritzen, Johannes and Hornig, Andreas and Winkler, Peter and Gude, Maik}}, booktitle = {{ECCM21 - Proceedings of the 21st European Conference on Composite Materials}}, isbn = {{978-2-912985-01-9}}, keywords = {{Direct parameter identification, Machine learning, Convolutional neural networks, Strain rate dependency, Fiber reinforced plastics, woven composites, segmentation, synthetic training data, x-ray computed tomography}}, pages = {{1252–1259}}, publisher = {{European Society for Composite Materials (ESCM)}}, title = {{{Direct parameter identification for highly nonlinear strain rate dependent constitutive models using machine learning}}}, doi = {{10.60691/yj56-np80}}, volume = {{3}}, year = {{2024}}, } @inproceedings{47522, abstract = {{Artificial benchmark functions are commonly used in optimization research because of their ability to rapidly evaluate potential solutions, making them a preferred substitute for real-world problems. However, these benchmark functions have faced criticism for their limited resemblance to real-world problems. In response, recent research has focused on automatically generating new benchmark functions for areas where established test suites are inadequate. These approaches have limitations, such as the difficulty of generating new benchmark functions that exhibit exploratory landscape analysis (ELA) features beyond those of existing benchmarks.The objective of this work is to develop a method for generating benchmark functions for single-objective continuous optimization with user-specified structural properties. Specifically, we aim to demonstrate a proof of concept for a method that uses an ELA feature vector to specify these properties in advance. To achieve this, we begin by generating a random sample of decision space variables and objective values. We then adjust the objective values using CMA-ES until the corresponding features of our new problem match the predefined ELA features within a specified threshold. By iteratively transforming the landscape in this way, we ensure that the resulting function exhibits the desired properties. To create the final function, we use the resulting point cloud as training data for a simple neural network that produces a function exhibiting the target ELA features. We demonstrate the effectiveness of this approach by replicating the existing functions of the well-known BBOB suite and creating new functions with ELA feature values that are not present in BBOB.}}, author = {{Prager, Raphael Patrick and Dietrich, Konstantin and Schneider, Lennart and Schäpermeier, Lennart and Bischl, Bernd and Kerschke, Pascal and Trautmann, Heike and Mersmann, Olaf}}, booktitle = {{Proceedings of the 17th ACM/SIGEVO Conference on Foundations of Genetic Algorithms}}, isbn = {{9798400702020}}, keywords = {{Benchmarking, Instance Generator, Black-Box Continuous Optimization, Exploratory Landscape Analysis, Neural Networks}}, pages = {{129–139}}, publisher = {{Association for Computing Machinery}}, title = {{{Neural Networks as Black-Box Benchmark Functions Optimized for Exploratory Landscape Features}}}, doi = {{10.1145/3594805.3607136}}, year = {{2023}}, } @inproceedings{52865, abstract = {{This paper addresses new challenges of detecting campaigns in social media, which emerged with the rise of Large Language Models (LLMs). LLMs particularly challenge algorithms focused on the temporal analysis of topical clusters. Simple similarity measures can no longer capture and map campaigns that were previously broadly similar in content. Herein, we analyze whether the classification of messages over time can be profitably used to rediscover poorly detectable campaigns at the content level. Thus, we evaluate classical classifiers and a new method based on siamese neural networks. Our results show that campaigns can be detected despite the limited reliability of the classifiers as long as they are based on a large amount of simultaneously spread artificial content.}}, author = {{Grimme, Britta and Pohl, Janina and Winkelmann, Hendrik and Stampe, Lucas and Grimme, Christian}}, booktitle = {{Disinformation in Open Online Media: 5th Multidisciplinary International Symposium, MISDOOM 2023, Amsterdam, The Netherlands, November 21–22, 2023, Proceedings}}, isbn = {{978-3-031-47895-6}}, keywords = {{Social Media, Campaign Detection, Large Language Models, Siamese Neural Networks}}, pages = {{72–87}}, publisher = {{Springer-Verlag}}, title = {{{Lost in Transformation: Rediscovering LLM-Generated Campaigns in Social Media}}}, doi = {{10.1007/978-3-031-47896-3_6}}, year = {{2023}}, } @inproceedings{33957, abstract = {{Manufacturing companies are challenged to make the increasingly complex work processes equally manageable for all employees to prevent an impending loss of competence. In this contribution, an intelligent assistance system is proposed enabling employees to help themselves in the workplace and provide them with competence-related support. This results in increasing the short- and long-term efficiency of problem solving in companies.}}, author = {{Deppe, Sahar and Brandt, Lukas and Brünninghaus, Marc and Papenkordt, Jörg and Heindorf, Stefan and Tschirner-Vinke, Gudrun}}, keywords = {{Assistance system, Knowledge graph, Information retrieval, Neural networks, AR}}, location = {{Stuttgart}}, title = {{{AI-Based Assistance System for Manufacturing}}}, doi = {{10.1109/ETFA52439.2022.9921520}}, year = {{2022}}, } @inproceedings{26539, abstract = {{In control design most control strategies are model-based and require accurate models to be applied successfully. Due to simplifications and the model-reality-gap physics-derived models frequently exhibit deviations from real-world-systems. Likewise, purely data-driven methods often do not generalise well enough and may violate physical laws. Recently Physics-Guided Neural Networks (PGNN) and physics-inspired loss functions separately have shown promising results to conquer these drawbacks. In this contribution we extend existing methods towards the identification of non-autonomous systems and propose a combined approach PGNN-L, which uses a PGNN and a physics-inspired loss term (-L) to successfully identify the system's dynamics, while maintaining the consistency with physical laws. The proposed method is demonstrated on two real-world nonlinear systems and outperforms existing techniques regarding complexity and reliability.}}, author = {{Götte, Ricarda-Samantha and Timmermann, Julia}}, booktitle = {{2022 3rd International Conference on Artificial Intelligence, Robotics and Control (AIRC)}}, keywords = {{data-driven, physics-based, physics-informed, neural networks, system identification, hybrid modelling}}, location = {{Cairo, Egypt}}, pages = {{67--76}}, title = {{{Composed Physics- and Data-driven System Identification for Non-autonomous Systems in Control Engineering}}}, doi = {{10.1109/AIRC56195.2022.9836982}}, year = {{2022}}, } @inproceedings{31066, abstract = {{While trade-offs between modeling effort and model accuracy remain a major concern with system identification, resorting to data-driven methods often leads to a complete disregard for physical plausibility. To address this issue, we propose a physics-guided hybrid approach for modeling non-autonomous systems under control. Starting from a traditional physics-based model, this is extended by a recurrent neural network and trained using a sophisticated multi-objective strategy yielding physically plausible models. While purely data-driven methods fail to produce satisfying results, experiments conducted on real data reveal substantial accuracy improvements by our approach compared to a physics-based model. }}, author = {{Schön, Oliver and Götte, Ricarda-Samantha and Timmermann, Julia}}, booktitle = {{14th IFAC Workshop on Adaptive and Learning Control Systems (ALCOS 2022)}}, keywords = {{neural networks, physics-guided, data-driven, multi-objective optimization, system identification, machine learning, dynamical systems}}, location = {{Casablanca, Morocco}}, number = {{12}}, pages = {{19--24}}, title = {{{Multi-Objective Physics-Guided Recurrent Neural Networks for Identifying Non-Autonomous Dynamical Systems}}}, doi = {{https://doi.org/10.1016/j.ifacol.2022.07.282}}, volume = {{55}}, year = {{2022}}, } @inproceedings{22480, abstract = {{In this publication important aspects for the implementation of inductive locating are explained. The miniaturized sensor platform called Sens-o-Spheres is used as an application of this locating method. The sensor platform is applied in bioreactors in order to obtain the environmental parameters, which makes a localization by magnetic fields necessary. Since the properties of magnetic fields in the localization area are very different from the wave characteristics, the principle of inductive localization is investigated in this publication and explained by using electrical equivalent circuit diagrams. Thereby, inductive localization uses the coupling or the mutual inductivities between coils, which is noticeable by an induced voltage. Therefore some properties and procedures are explained to extract the location of Sens-o-Spheres or other industrial sensor platforms from the couplings of the coils. One method calculates the location from an adapted ratio calculation and the other method uses neural networks and stochastic filters to obtain the results. In the end, these results are evaluated and compared.}}, author = {{Lange, Sven and Schröder, Dominik and Hedayat, Christian and Kuhn, Harald and Hilleringmann, Ulrich}}, booktitle = {{22nd IEEE International Conference on Industrial Technology (ICIT)}}, isbn = {{9781728157306}}, keywords = {{Location awareness, Coils, Couplings, Nonuniform electric fields, Magnetic separation, Neural networks, Training data}}, location = {{Valencia, Spain }}, publisher = {{IEEE}}, title = {{{Development of Methods for Coil-Based Localization by Magnetic Fields of Miniaturized Sensor Platforms in Bioprocesses}}}, doi = {{10.1109/icit46573.2021.9453609}}, year = {{2021}}, } @inproceedings{22481, abstract = {{During the industrial processing of materials for the manufacture of new products, surface defects can quickly occur. In order to achieve high quality without a long time delay, it makes sense to inspect the work pieces so that defective work pieces can be sorted out right at the beginning of the process. At the same time, the evaluation unit should come close the perception of the human eye regarding detection of defects in surfaces. Such defects often manifest themselves by a deviation of the existing structure. The only restriction should be that only matt surfaces should be considered here. Therefore in this work, different classification and image processing algorithms are applied to surface data to identify possible surface damages. For this purpose, the Gabor filter and the FST (Fused Structure and Texture) features generated with it, as well as the salience metric are used on the image processing side. On the classification side, however, deep neural networks, Convolutional Neural Networks (CNN), and autoencoders are used to make a decision. A distinction is also made between training using class labels and without. It turns out later that the salience metric are best performed by CNN. On the other hand, if there is no labeled training data available, a novelty classification can easily be achieved by using autoencoders as well as the salience metric and some filters.}}, author = {{Sander, Tom and Lange, Sven and Hilleringmann, Ulrich and Geneis, Volker and Hedayat, Christian and Kuhn, Harald and Gockel, Franz-Barthold}}, booktitle = {{22nd IEEE International Conference on Industrial Technology (ICIT)}}, isbn = {{9781728157306}}, keywords = {{Image Processing, Defect Detection, wooden surfaces, Machine Learning, Neural Networks}}, location = {{Valencia, Spain }}, publisher = {{IEEE}}, title = {{{Detection of Defects on Irregular Structured Surfaces by Image Processing Methods for Feature Extraction}}}, doi = {{10.1109/icit46573.2021.9453646}}, year = {{2021}}, } @inproceedings{27381, abstract = {{Graph neural networks (GNNs) have been successfully applied in many structured data domains, with applications ranging from molecular property prediction to the analysis of social networks. Motivated by the broad applicability of GNNs, we propose the family of so-called RankGNNs, a combination of neural Learning to Rank (LtR) methods and GNNs. RankGNNs are trained with a set of pair-wise preferences between graphs, suggesting that one of them is preferred over the other. One practical application of this problem is drug screening, where an expert wants to find the most promising molecules in a large collection of drug candidates. We empirically demonstrate that our proposed pair-wise RankGNN approach either significantly outperforms or at least matches the ranking performance of the naive point-wise baseline approach, in which the LtR problem is solved via GNN-based graph regression.}}, author = {{Damke, Clemens and Hüllermeier, Eyke}}, booktitle = {{Proceedings of The 24th International Conference on Discovery Science (DS 2021)}}, editor = {{Soares, Carlos and Torgo, Luis}}, isbn = {{9783030889418}}, issn = {{0302-9743}}, keywords = {{Graph-structured data, Graph neural networks, Preference learning, Learning to rank}}, location = {{Halifax, Canada}}, pages = {{166--180}}, publisher = {{Springer}}, title = {{{Ranking Structured Objects with Graph Neural Networks}}}, doi = {{10.1007/978-3-030-88942-5}}, volume = {{12986}}, year = {{2021}}, } @inproceedings{19953, abstract = {{Current GNN architectures use a vertex neighborhood aggregation scheme, which limits their discriminative power to that of the 1-dimensional Weisfeiler-Lehman (WL) graph isomorphism test. Here, we propose a novel graph convolution operator that is based on the 2-dimensional WL test. We formally show that the resulting 2-WL-GNN architecture is more discriminative than existing GNN approaches. This theoretical result is complemented by experimental studies using synthetic and real data. On multiple common graph classification benchmarks, we demonstrate that the proposed model is competitive with state-of-the-art graph kernels and GNNs.}}, author = {{Damke, Clemens and Melnikov, Vitaly and Hüllermeier, Eyke}}, booktitle = {{Proceedings of the 12th Asian Conference on Machine Learning (ACML 2020)}}, editor = {{Jialin Pan, Sinno and Sugiyama, Masashi}}, keywords = {{graph neural networks, Weisfeiler-Lehman test, cycle detection}}, location = {{Bangkok, Thailand}}, pages = {{49--64}}, publisher = {{PMLR}}, title = {{{A Novel Higher-order Weisfeiler-Lehman Graph Convolution}}}, volume = {{129}}, year = {{2020}}, } @inproceedings{15488, abstract = {{The continuous refinement of sensor technologies enables the manufacturing industry to capture increasing amounts of data during the production process. As processes take time to complete, sensors register large amounts of time-series-like data for each product. In order to make this data usable, a feature extraction is mandatory. In this work, we discuss and evaluate different network architectures, input pre-processing and cost functions regarding, among other aspects, their suitability for time series of different lengths.}}, author = {{Thiel, Christian and Steidl, Carolin and Henning, Bernd}}, booktitle = {{20. GMA/ITG-Fachtagung. Sensoren und Messsysteme 2019}}, isbn = {{978-3-9819376-0-2}}, keywords = {{Dynamic Time Warping, Feature Extraction, Masking, Neural Networks}}, title = {{{P2.9 Comparison of deep feature extraction techniques for varying-length time series from an industrial piercing press}}}, doi = {{10.5162/SENSOREN2019/P2.9}}, year = {{2019}}, }