Abstract
Currently, distributed software systems have evolved at an unprecedented pace. Modern software-quality requirements are high and require significant staff support and effort. This study investigates the use of a supervised machine learning model, a Multi-Layer Perceptron (MLP), for anomaly detection in microservices. The study covers the creation of a microservices infrastructure, the development of a fault injection module that simulates application-level and service-level anomalies, the creation of a system monitoring dataset, and the creation and validation of the MLP model to detect anomalies. The results indicate that the MLP model effectively detects anomalies in both domains with higher accuracy, precision, recovery, and F1 score on the service-level anomaly dataset. The potential for more effective distributed system monitoring and management automation is highlighted in this study by focusing on service-level metrics such as service response times. This study provides valuable information about the effectiveness of supervised machine learning models in detecting anomalies across distributed software systems.

Abstract

Abstract

Abstract

Abstract

Abstract
The human-environment-vehicle triad and how it relates to crashes has long been a topic of discussion, in which the human factor is consistently seen as the leading cause. Recently, more sophisticated approaches to Road Safety have advocated for a road-driver interaction view, in which human characteristics influence road perception and road environment affects driver behavior. This study focuses on road-driver interaction by using a driving simulator. The objective is to investigate how the driver profile influences driving performance and the effects of three countermeasures (peripheral transverse lines before and after the beginning of the curves and roadside poles in the curves). Fifty-six middle-aged male participants drove a non-challenging rural highway simulated scenario based on a real road where many single-vehicle crashes occurred. The drivers' profiles were assessed through their behavioral history measured by a validated version of the Driver Behavior Questionnaire (DBQ) comprising three dimensions: Errors (E), Ordinary Violations (OV), and Aggressive Violations (AV). The relationship between speed and trajectory measures and drivers' profiles was investigated using randomparameter models with heterogeneity in the means. The models' results showed that the DBQ subscale scores in OV explained a considerable part of the heterogeneity found in drivers' performance. Furthermore, the heterogeneity in the means caused by the DBQ subscale scores in OV and E in the presence of peripheral transverse lines indicates a difference in how drivers react to the countermeasures. The peripheral lines were more efficient than roadside poles to moderate speed but did not positively influence all drivers' trajectories. Although the peripheral lines could be seen as an alternative to change driver behavior in a non-challenging or monotonous road environment, the design used in this study should be reviewed.