Progress in making neural networks more robust against adversarial attacks is mostly marginal, despite the great efforts of the research community. Moreover, the robustness evaluation is often imprecise, making it challenging to identify promising approaches. We do an observational study on the classification decisions of 19 different state-of-the-art neural networks trained to be robust against adversarial attacks. This analysis gives a new indication of the limits of the robustness of current models on a common benchmark. In addition, our findings suggest that current untargeted adversarial attacks induce misclassification toward only a limited amount of different classes. Similarly, we find that previous attacks under-explore the perturbation space during optimization. This leads to unsuccessful attacks for samples where the initial gradient direction is not a good approximation of the final adversarial perturbation direction. Additionally, we observe that both over- and under-confidence in model predictions result in an inaccurate assessment of model robustness. Based on these observations, we propose a novel loss function for adversarial attacks that consistently improves their efficiency and success rate compared to prior attacks for all 30 analyzed models.

The most important goal of customer service is to keep the customer satisfied. However, service resources are always limited and must prioritize specific customers. Therefore, it is essential to identify customers who potentially become unsatisfied and might lead to escalations . Data science on IoT data (especially log data) for machine health monitoring and analytics on enterprise data for customer relationship management (CRM) have mainly been researched and applied independently. This paper presents a data-driven decision support system framework that combines IoT and enterprise data to model customer sentiment and predicts escalations. The proposed framework includes a fully automated and interpretable machine learning pipeline using state-of-the-art methods. The framework is applied in a real-world case study with a major medical device manufacturer providing data from a fleet of thousands of high-end medical devices. An anonymized version of this industrial benchmark is released for the research community based on the presented case study, which has interesting and challenging properties. In our extensive experiments, we achieve a Recall@50 of 50.0 % for the task of predicting customer escalations. In addition, we show that combining IoT and enterprise data can improve prediction results and ease troubleshooting. Additionally, we propose a practical workflow for end-users when applying the proposed framework.

Over the past few decades, the topic of artificial intelligence (AI) has gained considerable attention in both research and industry. In particular, the healthcare sector has witnessed a surge in the use of AI applications, as the maturity of these methods increased. However, as the use of machine learning (ML) in healthcare continues to grow, we believe it will become increasingly important to examine public perceptions of this trend to identify potential impediments and future directions. Current work focuses mainly on academic data sources and industrial applications of AI. However, to gain a comprehensive understanding of the increased societal interest in AI, digital media such as podcasts should be consulted, as they are accessible to a broader audience. In order to examine this hypothesis, we investigate the AI trend development in healthcare from 2015 until 2021. In this study, we propose a web mining approach to collect a novel data set consisting of 29 healthcare podcasts with 3449 episodes. We identify 102 AI-related buzzwords that were extracted from various glossaries and hype cycles. These buzzwords were used to conduct an extensive trend detection and analysis study on the collected data using machine learning-based approaches. We successfully detect an AI trend and follow its evolution in healthcare podcasts over several years. Besides the focus area of AI, we are able to detect 14 topic clusters and visualize the trending or decreasing dominant topics over the whole period under consideration. In addition, we analyze the sentiments in podcasts towards the identified topics and deliver further insights for trend detection in healthcare. Finally, the collected data set can be used for trend detection besides AI-related topics using topic clustering.

In mobile robotics, perception in uncontrolled environments like autonomous driving is a central hurdle. Existing active learning frameworks can help enhance perception by efficiently selecting data samples for labeling, but they are often constrained by the necessity of full data availability in data centers, hindering real-time, on-field adaptations. To address this, our work unveils a novel active learning formulation optimized for multi-robot settings. It harnesses the collaborative power of several robotic agents, considerably enhancing the data acquisition and synchronization processes. Experimental evidence indicates that our approach markedly surpasses traditional active learning frameworks by up to 2.5 percent points and 90% less data uploads, delivering new possibilities for advancements in the realms of mobile robotics and autonomous systems.

With tracking systems becoming more widespread in sports research and regular training and competitions, more data are available for sports analytics and performance prediction. We analyzed 2523 ski jumps from 205 athletes on five venues. For every jump, the dataset includes the 3D trajectory, 3D velocity, skis’ orientation, and metadata such as wind, starting gate, and ski jumping hill data. Using this dataset, we aimed to predict the expected jump length (xLength) inspired by the expected goals metric in soccer (xG). We evaluate the performance of a fully connected neural network, a convolutional neural network (CNN), a long short-term memory (LSTM), and a ResNet architecture to estimate the xLength. For the prediction of the jump length one second after take-off, we achieve a mean absolute error (MAE) of 5.3 $\mathrm{m}$ for the generalization to new athletes and an MAE of 5.9 $\mathrm{m}$ for the generalization to new ski jumping hills using ResNet architectures. Additionally, we investigated the influence of the input time after the take-off on the predictions’ accuracy. As expected, the MAE becomes smaller with longer inputs. Due to the real-time transmission of the sensor’s data, xLength can be updated during the flight phase and used in live TV broadcasting. xLength could also be used as an analysis tool for experts to quantify the quality of the take-off and flight phases.

Current research on trend detection in artificial intelligence (AI) mainly concerns academic data sources and industrial applications of AI. However, we argue that industrial trends are influenced by public perception and political decisions (e.g., through industry subsidies and grants) and should be reflected in political data sources. To investigate this hypothesis, we examine the AI trend development in German business and politics from 1998 to 2020. Therefore, we propose a web mining approach to collect a novel data set consisting of business and political data sources combining 1.07 million articles and documents. We identify 246 AI-related buzzwords extracted from various glossaries. We use them to conduct an extensive trend detection and analysis study on the collected data using machine learning-based approaches. This study successfully detects an AI trend and follows its evolution in business and political data sources over the past two decades. Moreover, we find a faster adoption of AI in business than in politics, with a considerable increase in policy discourse in recent years. Finally, we show that the collected data can be used for trend detection besides AI-related topics using topic clustering and the COVID-19 pandemic as examples.