A Trainable Centrality Framework for Modern Data
This work addresses the challenge of robust centrality estimation for modern data, offering a flexible and efficient solution for applications like outlier detection, but it is incremental as it builds on existing neural and classical methods.
The paper tackles the problem of measuring centrality in high-dimensional and non-Euclidean data, introducing FUSE, a neural framework that combines global and local signals to compute centrality scores, achieving competitive performance on outlier detection benchmarks across diverse data types.
Measuring how central or typical a data point is underpins robust estimation, ranking, and outlier detection, but classical depth notions become expensive and unstable in high dimensions and are hard to extend beyond Euclidean data. We introduce Fused Unified centrality Score Estimation (FUSE), a neural centrality framework that operates on top of arbitrary representations. FUSE combines a global head, trained from pairwise distance-based comparisons to learn an anchor-free centrality score, with a local head, trained by denoising score matching to approximate a smoothed log-density potential. A single parameter between 0 and 1 interpolates between these calibrated signals, yielding depth-like centrality from different views via one forward pass. Across synthetic distributions, real images, time series, and text data, and standard outlier detection benchmarks, FUSE recovers meaningful classical ordering, reveals multi-scale geometric structures, and attains competitive performance with strong classical baselines while remaining simple and efficient.