This study presents a supervised anomaly detection (AD) approach for real-world transaction networks, emphasizing cross-network generalization and structural feature extraction.Two signed and weighted Bitcoin trading platforms, namely Bitcoin Alpha and Bitcoin OTC, are analyzed by modeling user interactions as directed graphs, where edge weights encode trust and distrust ratings.Fraudulent behavior is detected by leveraging a combination of five graph-based centrality metrics alongside topological information derived from clustering.These features are used to train machine learning classifiers, tasked with identifying suspicious nodes.We conducted cross-network experiments in which the model is trained on one dataset and tested on the other, thereby simulating real-world generalization scenarios without incorporating temporal data.A comprehensive evaluation reveals that the proposed approach outperforms prior methods reported in the literature.Specifically, this study demonstrates high detection capability, low false alarm rates, and strong overall robustness across both datasets.

We present a O(|A|(|V | + |A|)) heuristic Algorithm for the minimum weighted feedback arc set problem.The Algorithm is based on two linear arrangements of the set of arcs, and on the concept of minimal and stable feedback arc set previously introduced and now generalized to the weighted case.The obtained results, compared to two other state of the art Algorithms, show a high level of efficacy of the proposed Algorithm.

The growing availability of geo–referred data describing human behaviour, at different scales and levels of granularity, represents an opportunity for the development and application of data analysis algorithms, whose usage can range from security, to traffic, to architectural design and planning, and even marketing. Focusing on pedestrian generated trajectories, the presence of groups within an analyzed population can influence overall dynamics, from microscopic perspective, and it can provide significant indications. Several approaches for video footage analyses are available, but they generally focus on microscopic features of videos and trajectories and they are generally not suited to scale to the analysis of relatively large datasets of trajectories. The present work proposes a novel approach to spatial–temporal analysis of pedestrian trajectories aimed at detecting groups of pedestrians within large datasets and having minimal assumptions on the nature of these groups.