Refine
Document Type
- Conference Proceeding (12)
- Article (unreviewed) (5)
- Bachelor Thesis (3)
- Part of a Book (2)
- Master's Thesis (2)
- Article (reviewed) (1)
Conference Type
- Konferenzartikel (12)
Keywords
- Machine Learning (25) (remove)
Institute
- Fakultät Elektrotechnik, Medizintechnik und Informatik (EMI) (ab 04/2019) (15)
- IMLA - Institute for Machine Learning and Analytics (9)
- Fakultät Medien (M) (ab 22.04.2021) (4)
- Fakultät Wirtschaft (W) (3)
- Fakultät Maschinenbau und Verfahrenstechnik (M+V) (2)
- IfTI - Institute for Trade and Innovation (2)
- Fakultät Elektrotechnik und Informationstechnik (E+I) (bis 03/2019) (1)
- INES - Institut für nachhaltige Energiesysteme (1)
Open Access
- Open Access (14)
- Closed (7)
- Bronze (4)
- Closed Access (3)
- Diamond (1)
Soiling is an important issue in the renewable energy sector since it can result in significant yield losses, especially in regions with higher pollution or dust levels. To mitigate the impact of soiling on photovoltaic (PV) plants, it is essential to regularly monitor and clean the panels, as well as develop accurate soiling predictions that can affect cleaning strategies and enhance the overall performance of PV power plants. This research focuses on the problem of soiling loss in photovoltaic power plants and the potential to improve the accuracy of soiling predictions. The study examines how soiling can affect the efficiency and productivity of the modules and how to measure and predict soiling using machine learning (ML) algorithms. The research includes analyzing real data from large-scale ground-mounted PV sites and comparing different soiling measurement methods. It was observed that there were some deviations in the real soiling loss values compared to the expected values for some projects in southern Spain, thus, the main goal of this work is to develop machine learning models that could predict the soiling more accurately. The developed models have a low mean square error (MSE), indicating the accuracy and suitability of the models to predict the soiling rates. The study also investigates the impact of different cleaning strategies on the performance of PV power plants and provides a powerful application to predict both the soiling and the number of cleaning cycles.
The progress in machine learning has led to advanced deep neural networks. These networks are widely used in computer vision tasks and safety-critical applications. The automotive industry, in particular, has experienced a significant transformation with the integration of deep learning techniques and neural networks. This integration contributes to the realization of autonomous driving systems. Object detection is a crucial element in autonomous driving. It contributes to vehicular safety and operational efficiency. This technology allows vehicles to perceive and identify their surroundings. It detects objects like pedestrians, vehicles, road signs, and obstacles. Object detection has evolved from being a conceptual necessity to an integral part of advanced driver assistance systems (ADAS) and the foundation of autonomous driving technologies. These advancements enable vehicles to make real-time decisions based on their understanding of the environment, improving safety and driving experiences. However, the increasing reliance on deep neural networks for object detection and autonomous driving has brought attention to potential vulnerabilities within these systems. Recent research has highlighted the susceptibility of these systems to adversarial attacks. Adversarial attacks are well-designed inputs that exploit weaknesses in the deep learning models underlying object detection. Successful attacks can cause misclassifications and critical errors, posing a significant threat to the functionality and safety of autonomous vehicles. With the rapid development of object detection systems, the vulnerability to adversarial attacks has become a major concern. These attacks manipulate inputs to deceive the target system, significantly compromising the reliability and safety of autonomous vehicles. In this study, we focus on analyzing adversarial attacks on state-of-the-art object detection models. We create adversarial examples to test the models’ robustness. We also check if the attacks work on a different object detection model meant for similar tasks. Additionally, we extensively evaluate recent defense mechanisms to see how effective they are in protecting deep neural networks (DNNs) from adversarial attacks and provide a comprehensive overview of the most commonly used defense strategies against adversarial attacks, highlighting how they can be implemented practically in real-world situations.
Prediction of Claims in Export Credit Finance: A Comparison of Four Machine Learning Techniques
(2020)
This study evaluates four machine learning (ML) techniques (Decision Trees (DT), Random Forests (RF), Neural Networks (NN) and Probabilistic Neural Networks (PNN)) on their ability to accurately predict export credit insurance claims. Additionally, we compare the performance of the ML techniques against a simple benchmark (BM) heuristic. The analysis is based on the utilisation of a dataset provided by the Berne Union, which is the most comprehensive collection of export credit insurance data and has been used in only two scientific studies so far. All ML techniques performed relatively well in predicting whether or not claims would be incurred, and, with limitations, in predicting the order of magnitude of the claims. No satisfactory results were achieved predicting actual claim ratios. RF performed significantly better than DT, NN and PNN against all prediction tasks, and most reliably carried their validation performance forward to test performance.
Polyarticulated active prostheses constitute a promising solution for upper limb amputees. The bottleneck for their adoption though, is the lack of intuitive control. In this context, machine learning algorithms based on pattern recognition from electromyographic (EMG) signals represent a great opportunity for naturally operating prosthetic devices, but their performance is strongly affected by the selection of input features. In this study, we investigated different combinations of 13 EMG-derived features obtained from EMG signals of healthy individuals performing upper limb movements and tested their performance for movement classification using an Artificial Neural Network. We found that input data (i.e., the set of input features) can be reduced by more than 50% without any loss in accuracy, while diminishing the computing time required to train the classifier. Our results indicate that input features must be properly selected in order to optimize prosthetic control.
In this paper we show that a model-free approach to learn behaviors in joint space can be successfully used to utilize toes of a humanoid robot. Keeping the approach model-free makes it applicable to any kind of humanoid robot, or robot in general. Here we focus on the benefit on robots with toes which is otherwise more difficult to exploit. The task has been to learn different kick behaviors on simulated Nao robots with toes in the RoboCup 3D soccer simulator. As a result, the robot learned to step on its toe for a kick that performs 30% better than learning the same kick without toes.
A fundamental and still largely unsolved question in the context of Generative Adversarial Networks is whether they are truly able to capture the real data distribution and, consequently, to sample from it. In particular, the multidimensional nature of image distributions leads to a complex evaluation of the diversity of GAN distributions. Existing approaches provide only a partial understanding of this issue, leaving the question unanswered. In this work, we introduce a loop-training scheme for the systematic investigation of observable shifts between the distributions of real training data and GAN generated data. Additionally, we introduce several bounded measures for distribution shifts, which are both easy to compute and to interpret. Overall, the combination of these methods allows an explorative investigation of innate limitations of current GAN algorithms. Our experiments on different data-sets and multiple state-of-the-art GAN architectures show large shifts between input and output distributions, showing that existing theoretical guarantees towards the convergence of output distributions appear not to be holding in practice.
Deep generative models have recently achieved impressive results for many real-world applications, successfully generating high-resolution and diverse samples from complex datasets. Due to this improvement, fake digital contents have proliferated growing concern and spreading distrust in image content, leading to an urgent need for automated ways to detect these AI-generated fake images.
Despite the fact that many face editing algorithms seem to produce realistic human faces, upon closer examination, they do exhibit artifacts in certain domains which are often hidden to the naked eye. In this work, we present a simple way to detect such fake face images - so-called DeepFakes. Our method is based on a classical frequency domain analysis followed by basic classifier. Compared to previous systems, which need to be fed with large amounts of labeled data, our approach showed very good results using only a few annotated training samples and even achieved good accuracies in fully unsupervised scenarios. For the evaluation on high resolution face images, we combined several public datasets of real and fake faces into a new benchmark: Faces-HQ. Given such high-resolution images, our approach reaches a perfect classification accuracy of 100% when it is trained on as little as 20 annotated samples. In a second experiment, in the evaluation of the medium-resolution images of the CelebA dataset, our method achieves 100% accuracy supervised and 96% in an unsupervised setting. Finally, evaluating a low-resolution video sequences of the FaceForensics++ dataset, our method achieves 91% accuracy detecting manipulated videos.
In this preliminary report, we present a simple but very effective technique to stabilize the training of CNN based GANs. Motivated by recently published methods using frequency decomposition of convolutions (e.g. Octave Convolutions), we propose a novel convolution scheme to stabilize the training and reduce the likelihood of a mode collapse. The basic idea of our approach is to split convolutional filters into additive high and low frequency parts, while shifting weight updates from low to high during the training. Intuitively, this method forces GANs to learn low frequency coarse image structures before descending into fine (high frequency) details. Our approach is orthogonal and complementary to existing stabilization methods and can simply plugged into any CNN based GAN architecture. First experiments on the CelebA dataset show the effectiveness of the proposed method.
Recent work has investigated the distributions of learned convolution filters through a large-scale study containing hundreds of heterogeneous image models. Surprisingly, on average, the distributions only show minor drifts in comparisons of various studied dimensions including the learned task, image domain, or dataset. However, among the studied image domains, medical imaging models appeared to show significant outliers through "spikey" distributions, and, therefore, learn clusters of highly specific filters different from other domains. Following this observation, we study the collected medical imaging models in more detail. We show that instead of fundamental differences, the outliers are due to specific processing in some architectures. Quite the contrary, for standardized architectures, we find that models trained on medical data do not significantly differ in their filter distributions from similar architectures trained on data from other domains. Our conclusions reinforce previous hypotheses stating that pre-training of imaging models can be done with any kind of diverse image data.
Many commonly well-performing convolutional neural network models have shown to be susceptible to input data perturbations, indicating a low model robustness. Adversarial attacks are thereby specifically optimized to reveal model weaknesses, by generating small, barely perceivable image perturbations that flip the model prediction. Robustness against attacks can be gained for example by using adversarial examples during training, which effectively reduces the measurable model attackability. In contrast, research on analyzing the source of a model’s vulnerability is scarce. In this paper, we analyze adversarially trained, robust models in the context of a specifically suspicious network operation, the downsampling layer, and provide evidence that robust models have learned to downsample more accurately and suffer significantly less from aliasing than baseline models.
Jeder Mensch ist ständig unfreiwillig von einer Flut akustischer Reize umgeben. Diese Situation stellt für Menschen mit Hörverlust eine besondere Herausforderung dar. Menschen mit Hörverlust hören durch Hörgeräte zwar alles verstärkt, jedoch stellt sich die Frage, ob ein Hörgerät lediglich eine einfache Verstärkung von Schallwellen ist oder ob es darüber hinausgehende Funktionen bieten kann.
Die vorliegende Thesis widmet sich der akustischen Szenenanalyse in Hörgeräten, wobei der Schwerpunkt auf der Integration von Machine Learning liegt. Das Ziel besteht darin, eine automatisierte Erkennung und Anpassung an verschiedene akustische Situationen zu ermöglichen. Die Arbeit konzentriert sich insbesondere auf die Analyse grundlegender Szenarien wie: Sprache in Ruhe, absolute Ruhe, Sprache in Störgeräuschen und Störgeräuschen in Audiodaten.
The present work ties in with the problem of bicycle road assessment that is currently done using expensive special measuring vehicles. Our alternative approach for road condition assessment is to mount a sensor device on a bicycle which sends accelerometer and gyroscope data via WiFi to a classification server. There, a prediction model determines road type and condition based on the sensor data. For the classification task, we compare different machine learning methods with each other, whereby validation accuracies of 99% can be achieved with deep residual networks such as InceptionTime. The main contribution of this work with respect to comparable work is that we achieve excellent accuracies on a realistic dataset classifying road conditions into nine distinct classes that are highly relevant for practice.
Estimating the Robustness of Classification Models by the Structure of the Learned Feature-Space
(2022)
Over the last decade, the development of deep image classification networks has mostly been driven by the search for the best performance in terms of classification accuracy on standardized benchmarks like ImageNet. More recently, this focus has been expanded by the notion of model robustness, \ie the generalization abilities of models towards previously unseen changes in the data distribution. While new benchmarks, like ImageNet-C, have been introduced to measure robustness properties, we argue that fixed testsets are only able to capture a small portion of possible data variations and are thus limited and prone to generate new overfitted solutions. To overcome these drawbacks, we suggest to estimate the robustness of a model directly from the structure of its learned feature-space. We introduce robustness indicators which are obtained via unsupervised clustering of latent representations from a trained classifier and show very high correlations to the model performance on corrupted test data.
In order to make material design processes more efficient in the future, the underlying multidimensional process parameter spaces must be systematically explored using digitalisation techniques such as machine learning (ML) and digital simulation. In this paper we shortly review essential concepts for the digitalisation of electrodeposition processes with a special focus on chromium plating from trivalent electrolytes.
Diese Arbeit befasst sich mit dem Entwurf und der Herstellung einer Roboterhandprothese, die amputierten Menschen eine gewisse Mobilität und eine teilweise Nutzung der Hand ermöglichen soll.
Das Projekt konzentriert sich insbesondere auf die Erkennung der vom Benutzer ausgeführten Bewegung und wird die Schritte der Erfassung, der Bewegung der Übertragung und die Erkennung detailliert darstellen.
Recently, RobustBench (Croce et al. 2020) has become a widely recognized benchmark for the adversarial robustness of image
classification networks. In it’s most commonly reported sub-task, RobustBench evaluates and ranks the adversarial robustness of trained neural networks on CIFAR10 under AutoAttack (Croce and Hein 2020b) with l∞ perturbations limited to ϵ = 8/255. With leading scores of the currently best performing models of around 60% of the baseline, it is fair to characterize this benchmark to be quite challenging. Despite it’s general acceptance in recent literature, we aim to foster discussion about the suitability of RobustBench as a key indicator for robustness which could be generalized to practical applications. Our line of argumentation against this is two-fold and supported by excessive experiments presented in this paper: We argue that I) the alternation of data by AutoAttack with l∞, ϵ = 8/255 is unrealistically strong, resulting in close to perfect detection rates of adversarial samples even by simple detection algorithms and human observers.
We also show that other attack methods are much harder to detect while achieving similar success rates. II) That results on low resolution data sets like CIFAR10 do not generalize well to higher resolution images as gradient based attacks appear to become even more detectable with increasing resolutions.
Diese Bachelorthesis behandelt die Entwicklung eines Prototyps zur Identifizierung und Verhinderung von Angriffen mithilfe von KI- und ML-Modellen. Untersucht werden die Leistungsfähigkeit verschiedener theoretischer Modelle im Kontext der Intrusion Detection, wobei Machine-Learning-Modelle wie Entscheidungsbäume, Random Forests und Naive Bayes analysiert werden. Die Arbeit betont die Relevanz der Datensatzauswahl, die Vorbereitung der Daten und bietet einen Ausblick auf zukünftige Entwicklungen in der Angriffserkennung.
Harnessing the overall benefits of the latest advancements in artificial intelligence (AI) requires the extensive collaboration of academia and industry. These collaborations promote innovation and growth while enforcing the practical usefulness of newer technologies in real life. The purpose of this article is to outline the challenges faced during cross-collaboration between academia and industry. These challenges are also inspected with the help of an ongoing project titled “Quality Assurance of Machine Learning Applications” (Q-AMeLiA), in which three universities cooperate with five industry partners to make the product risk of AI-based products visible. Further, we discuss the hurdles and the key challenges in machine learning (ML) technology transformation from academia to industry based on robustness, simplicity, and safety. These challenges are an outcome of the lack of common standards, metrics, and missing regulatory considerations when state-of-the-art (SOTA) technology is developed in academia. The use of biased datasets involves ethical concerns that might lead to unfair outcomes when the ML model is deployed in production. The advancement of AI in small and medium sized enterprises (SMEs) requires more in terms of common tandardization of concepts rather than algorithm breakthroughs. In this paper, in addition to the general challenges, we also discuss domain specific barriers for five different domains i.e., object detection, hardware benchmarking, continual learning, action recognition, and industrial process automation, and highlight the steps necessary for successfully managing the cross-sectoral collaborations between academia and industry.
Apache Hadoop is a well-known open-source framework for storing and processing huge amounts of data. This paper shows the usage of the framework within a project of the university in cooperation with a semiconductor company. The goal of this project was to supplement the existing data landscape by the facilities of storing and analyzing the data on a new Apache Hadoop based platform.
Machine Learning (ML) als das aktuell wohl am meisten diskutierte Teilgebiet von Künstlicher Intelligenz (KI) verspricht und realisiert bereits sinnvolle Unterstützung u.a. beim autonomen Fahren, der Predictive Maintenance, der Verbrechensvorbeugung, dem Zusammenführen von Angebot und Nachfrage durch Empfehlungslisten oder im Kundenservice mit Softbots.1 All diesen Anwendung ist letztlich gemeinsam, dass Entscheidungen zu treffen sind, und zwar möglichst rational vor dem Hintergrund von subjektiven Präferenzsystemen und den in Betracht gezogenen Handlungsalternativen. Die Entscheidungsgegenstände betreffen Fragen, wann ein Fahrzeug leicht abgebremst werden soll, in welcher Gegend schwerpunktmäßig Polizeipräsenz gezeigt werden soll oder wann eine Maschine gewartet werden soll.
In the field of network security, the detection of possible intrusions is an important task to prevent and analyse attacks. Machine learning has been adopted as a particular supporting technique over the last years. However, the majority of related published work uses post mortem log files and fails to address the required real-time capabilities of network data feature extraction and machine learning based analysis [1-5]. We introduce the network feature extractor library FEX, which is designed to allow real-time feature extraction of network data. This library incorporates 83 statistical features based on reassembled data flows. The introduced Cython implementation allows processing individual packets within 4.58 microseconds. Based on the features extracted by FEX, existing intrusion detection machine learning models were examined with respect to their real-time capabilities. An identified Decision-Tree Classifier model was thus further optimised by transpiling it into C Code. This reduced the prediction time of a single sample to 3.96 microseconds on average. Based on the feature extractor and the improved machine learning model an IDS system was implemented which supports a data throughput between 63.7 Mbit/s and 2.5 Gbit/s making it a suitable candidate for a real-time, machine-learning based IDS.
As e-commerce platforms have grown in popularity, new difficulties have emerged, such as the growing use of bots—automated programs—to engage with e-commerce websites. Even though some algorithms are helpful, others are malicious and can seriously hurt e-commerce platforms by making fictitious purchases, posting fictitious evaluations, and gaining control of user accounts. Therefore, the development of more effective and precise bot identification systems is urgently needed to stop such actions. This thesis proposes a methodology for detecting bots in E-commerce using machine learning algorithms such as K-nearest neighbors, Decision Tree, Random Forest, Support Vector Machine, and Neural Network. The purpose of the research is to assess and contrast the output of these machine learning methods. The suggested approach will be based on data that is readily accessible to the public, and the study’s focus will be on the research of bots in e-commerce.
The purpose of the study is to provide an overview of bots in e-commerce, as well as information on the different kinds and traits of bots, as well as current research on bots in e-commerce and associated work on bot detection in e-commerce. The research also seeks to create a more precise and effective bot detection system as well as find critical factors in detecting bots in e-commerce.
This research is significant because it sheds light on the increasing issue of bots in e-commerce and the requirement for more effective bot detection systems. The suggested approach for using machine learning algorithms to identify bots in ecommerce can give e-commerce platforms a more precise and effective bot detection system to stop malicious bot activities. The study’s results can also be used to create a more effective bot detection system and pinpoint key elements in detecting bots in e-commerce.
Machine Learning als Schlüsseltechnologie für Digitalisierung: Wie funktioniert maschinelles Lernen?
(2019)
In diesem Beitrag werden grundlegende Aspekte und Methoden der Data Science erläutert. Nach dem Vorgehensmodell CRISP-DM sind in den Phasen Data Unterstanding und Data Preparation vor allem Verfahren der Datenselektion, Datenvorverarbeitung und der explorativen Datenanalyse anzuwenden. Beim Modeling, der Hauptaufgabe der Data Science, kann man überwachte und unüberwachte Methoden sowie Reinforcement Learning unterscheiden. Auf die Evaluation der Güte eines Modells anhand von Qualitätsmaßen wird anschließend eingegangen. Der Beitrag schließt mit einem Ausblick auf weitere Themen wie Cognitive Computing.