Active Neuron Academy
Insights & publications
Graph learning, dynamic networks, and AML: peer-reviewed work that informs Active Neuron Academy programmes (University of Porto, Feedzai collaboration, and related venues). Each card opens the paper (OpenReview, arXiv, or ACM).
Publications
Newest first
Deep-Graph-Sprints: accelerated representation learning in continuous-time dynamic graphs
Low-latency deep representation learning on continuous-time dynamic graphs.
Open paper
From random-walks to graph-sprints: low-latency node embeddings on continuous-time dynamic graphs
Streaming, low-latency approximation to random-walk features for CTDGs.
arXiv:2307.08433
Anti-money laundering alert optimization using machine learning with graphs
ML triage with entity and graph features on dynamic graphs; also related patent work.
arXiv:2112.07508
Scalable subgraph counting using MapReduce
MSc-era work; Motif-Discovery plugin reached 10k+ downloads.
ACM Digital Library