Beyond Distance: Quantifying Point Cloud Dynamics with Persistent Homology and Dynamic Optimal Transport
2026-03-18T07:23:57Z•4030cd4457c84e44d50831d83a0a5d491c617c380717d284b8fe34be72c046af
Lipschitz optimization (HALO)adaptive design of experimentsconditional distributional effectscovariate shiftdeep adaptive designdistributional robustnessdoubly robust estimationentropy metricsfeature selectionglobal optimizationhypergraphshypothesis testingk-NNnon-orthogonal embeddingsoptimal transportpartial-label learningpermutation-free testspersistent homologypoint cloudsrepresentation learningsafe screeningsparse sensingtopological data analysistransformer policytransformers
What happened
This collection of recent arXiv submissions spans methodological advances and theory across machine learning, statistics, optimization, quantum computing, and scientific decision support. Highlights include: a Topological Optimal Transport extension for time-evolving point clouds with multi-scale entropy and hypergraph reconstruction for detecting localized topological tipping; a provably effective adaptive k-NN method and feasibility conditions for partial-label learning; analysis of a single-layer transformer trained with random (non-orthogonal) embeddings giving explicit storage-capacity/sc
Why it matters
A reviewed impact interpretation has not been published for this record.
Evidence and limitations
- Source ID
- arxiv_stat_ml
- Record identifier
- 4030cd4457c84e44d50831d83a0a5d491c617c380717d284b8fe34be72c046af
- Enrichment time
- 2026-03-18T07:23:57Z
- AI-assisted enrichment
- Yes
This record may overlap with other records. Its enrichment can be incomplete or wrong, and machine assistance was used. Validate consequential decisions against the linked source and your own environment.