Termination analyses investigate the termination behavior of programs, intending to detect nontermination, which is known to cause a variety of program bugs (e.g. hanging programs, denial-of-service vulnerabilities). Beyond formal approaches, various attempts have been made to estimate the termination behavior of programs using neural networks. However, the majority of these approaches continue to rely on formal methods to provide strong soundness guarantees and consequently suffer from similar limitations. In this paper, we move away from formal methods and embrace the stochastic nature of machine learning models. Instead of aiming for rigorous guarantees that can be interpreted by solvers, our objective is to provide an estimation of a program's termination behavior and of the likely reason for nontermination (when applicable) that a programmer can use for debugging purposes. Compared to previous approaches using neural networks for program termination, we also take advantage of the graph representation of programs by employing Graph Neural Networks. To further assist programmers in understanding and debugging nontermination bugs, we adapt the notions of attention and semantic segmentation, previously used for other application domains, to programs. Overall, we designed and implemented classifiers for program termination based on Graph Convolutional Networks and Graph Attention Networks, as well as a semantic segmentation Graph Neural Network that localizes AST nodes likely to cause nontermination. We also illustrated how the information provided by semantic segmentation can be combined with program slicing to further aid debugging.

Machine learning and computer vision techniques have grown rapidly in recent years due to their automation, suitability, and ability to generate astounding results. Hence, in this paper, we survey the key studies that are published between 2014 and 2022, showcasing the different machine learning algorithms researchers have used to segment the liver, hepatic tumors, and hepatic-vasculature structures. We divide the surveyed studies based on the tissue of interest (hepatic-parenchyma, hepatic-tumors, or hepatic-vessels), highlighting the studies that tackle more than one task simultaneously. Additionally, the machine learning algorithms are classified as either supervised or unsupervised, and they are further partitioned if the amount of work that falls under a certain scheme is significant. Moreover, different datasets and challenges found in literature and websites containing masks of the aforementioned tissues are thoroughly discussed, highlighting the organizers' original contributions and those of other researchers. Also, the metrics used excessively in literature are mentioned in our review, stressing their relevance to the task at hand. Finally, critical challenges and future directions are emphasized for innovative researchers to tackle, exposing gaps that need addressing, such as the scarcity of many studies on the vessels' segmentation challenge and why their absence needs to be dealt with sooner than later.

Machine learning (ML) is increasingly being used to support high-stakes decisions, a trend owed in part to its promise of superior predictive power relative to human assessment. However, there is frequently a gap between decision objectives and what is captured in the observed outcomes used as labels to train ML models. As a result, machine learning models may fail to capture important dimensions of decision criteria, hampering their utility for decision support. In this work, we explore the use of historical expert decisions as a rich -- yet imperfect -- source of information that is commonly available in organizational information systems, and show that it can be leveraged to bridge the gap between decision objectives and algorithm objectives. We consider the problem of estimating expert consistency indirectly when each case in the data is assessed by a single expert, and propose influence function-based methodology as a solution to this problem. We then incorporate the estimated expert consistency into a predictive model through a training-time label amalgamation approach. This approach allows ML models to learn from experts when there is inferred expert consistency, and from observed labels otherwise. We also propose alternative ways of leveraging inferred consistency via hybrid and deferral models. In our empirical evaluation, focused on the context of child maltreatment hotline screenings, we show that (1) there are high-risk cases whose risk is considered by the experts but not wholly captured in the target labels used to train a deployed model, and (2) the proposed approach significantly improves precision for these cases.

Gold mining is one of the very oldest human occupations. The earliest known underground gold mine, in what is now the country of Georgia, dates back at least 5,000 years, when people were just starting to develop written language. Over the centuries, a number of innovations have emerged that disrupted and forever changed how we explore and mine for gold and other metals. Think dynamite, or the steam engine. Lately, however, innovation has slowed.

Wearable robotic devices have potential to assist and protect their users. Toward design of a Smart Helmet, this article examines the effectiveness of audio and visual warnings to help participants brace for impacts. A user study examines different warnings and impacts applied to users while running. Perturbation forces scaled to user mass are applied from different directions and user displacement is measured to characterize effectiveness of the warning. This is accomplished using the TreadPort Active Wind Tunnel adapted to deliver forward, rearward, right, or left perturbation forces at precise moments during the locomotor cycle. The article presents an overview of the system and demonstrates the ability to precisely deliver consistent warnings and perturbations during gait. User study results highlight effectiveness of visual and audio warnings to help users brace for impact, resulting in guidelines that will inform future human-robot warning systems.

Uncertainty quantification is one of the central challenges for machine learning in real-world applications. In reinforcement learning, an agent confronts two kinds of uncertainty, called epistemic uncertainty and aleatoric uncertainty. Disentangling and evaluating these uncertainties simultaneously stands a chance of improving the agent's final performance, accelerating training, and facilitating quality assurance after deployment. In this work, we propose an uncertainty-aware reinforcement learning algorithm for continuous control tasks that extends the Deep Deterministic Policy Gradient algorithm (DDPG). It exploits epistemic uncertainty to accelerate exploration and aleatoric uncertainty to learn a risk-sensitive policy. We conduct numerical experiments showing that our variant of DDPG outperforms vanilla DDPG without uncertainty estimation in benchmark tasks on robotic control and power-grid optimization.

Hyperparameter optimization constitutes a large part of typical modern machine learning workflows. This arises from the fact that machine learning methods and corresponding preprocessing steps often only yield optimal performance when hyperparameters are properly tuned. But in many applications, we are not only interested in optimizing ML pipelines solely for predictive accuracy; additional metrics or constraints must be considered when determining an optimal configuration, resulting in a multi-objective optimization problem. This is often neglected in practice, due to a lack of knowledge and readily available software implementations for multi-objective hyperparameter optimization. In this work, we introduce the reader to the basics of multi-objective hyperparameter optimization and motivate its usefulness in applied ML. Furthermore, we provide an extensive survey of existing optimization strategies, both from the domain of evolutionary algorithms and Bayesian optimization. We illustrate the utility of MOO in several specific ML applications, considering objectives such as operating conditions, prediction time, sparseness, fairness, interpretability and robustness.

In the recent years, social media has grown to become a major source of information for many online users. This has given rise to the spread of misinformation through deepfakes. Deepfakes are videos or images that replace one persons face with another computer-generated face, often a more recognizable person in society. With the recent advances in technology, a person with little technological experience can generate these videos. This enables them to mimic a power figure in society, such as a president or celebrity, creating the potential danger of spreading misinformation and other nefarious uses of deepfakes. To combat this online threat, researchers have developed models that are designed to detect deepfakes. This study looks at various deepfake detection models that use deep learning algorithms to combat this looming threat. This survey focuses on providing a comprehensive overview of the current state of deepfake detection models and the unique approaches many researchers take to solving this problem. The benefits, limitations, and suggestions for future work will be thoroughly discussed throughout this paper.

Indeed, commercial task-oriented dialog systems in the form of smart devices like Amazon's Alexa are used by millions of people every day. Within the academic research community, however, task-oriented dialog system models are often benchmarked on relatively few evaluation datasets. This is in spite of the fact that the past few years have seen a substantial growth in the number of available datasets for building and evaluating intent classification and slot-filling models for task-oriented dialog systems. Thus, the goal of this survey is to catalog these intent classification and slot-filling datasets to help facilitate their use in building and evaluating dialog systems and beyond. Other surveys have discussed dialog datasets in depth (Serban et al. 2018), but exclude almost all intent classification and slot-filling datasets, and model-focused surveys on dialog systems mostly focus on models and pay much less attention to datasets.

Small object detection (SOD) in optical images and videos is a challenging problem that even state-of-the-art generic object detection methods fail to accurately localize and identify such objects. Typically, small objects appear in real-world due to large camera-object distance. Because small objects occupy only a small area in the input image (e.g., less than 10%), the information extracted from such a small area is not always rich enough to support decision making. Multidisciplinary strategies are being developed by researchers working at the interface of deep learning and computer vision to enhance the performance of SOD deep learning based methods. In this paper, we provide a comprehensive review of over 160 research papers published between 2017 and 2022 in order to survey this growing subject. This paper summarizes the existing literature and provide a taxonomy that illustrates the broad picture of current research. We investigate how to improve the performance of small object detection in maritime environments, where increasing performance is critical. By establishing a connection between generic and maritime SOD research, future directions have been identified. In addition, the popular datasets that have been used for SOD for generic and maritime applications are discussed, and also well-known evaluation metrics for the state-of-the-art methods on some of the datasets are provided.