Research map / Label-efficient and robust learning / Semi-supervised learning
RSDNE: Exploring Relaxed Similarity and Dissimilarity from Completely-Imbalanced Labels for Network Embedding
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add intra-class similarity modeling (adaptive k-neighbor selection) to the DeepWalk structure model (ablation) | RSDNE · Citeseer, completely-imbalanced labels | Micro-F1/Macro-F1 | Helped |
| add inter-class dissimilarity modeling (remove links between differently labeled nodes) (ablation) | RSDNE · Citeseer, completely-imbalanced labels | Micro-F1/Macro-F1 | Helped |
| select intra-class neighbors adaptively instead of randomly (ablation) | RSDNE · Citeseer | Micro-F1 | Helped |
| add random intra-class neighbor selection over structure-only model (ablation) | DeepWalk matrix factorization (JDW) · Citeseer | Micro-F1 | Helped |
| use RSDNE instead of classifier-based semi-supervised methods (LSHM, LDE, MMDW) | semi-supervised network embedding with unseen classes · Citeseer, Cora, Wiki | Micro-F1: 7–12% relative over MMDW at 50% labels | Helped |
| restrict intra-class neighbor candidates to a random subset (light version RSDNE*) (ablation) | RSDNE · Citeseer, Cora, Wiki | Micro-F1 | No clear effect |
| add classifier-based labels (LSHM, LDE) when some classes are unseen vs unsupervised DeepWalk/MFDW | semi-supervised baselines · Citeseer, Cora, Wiki | accuracy | Hurt |
| use RSDNE under balanced labels instead of other semi-supervised methods | semi-supervised network embedding · balanced labels, three datasets | Micro-F1 | No clear effect |
Closest papers by meaning
- Conditional Network Embeddings (ICLR 2019 · similarity 0.90)
- Reconciling Competing Sampling Strategies of Network Embedding (NeurIPS 2023 · similarity 0.87)
- Revisiting Semi-Supervised Learning with Graph Embeddings (ICML 2016 · similarity 0.87)
- Structural Deep Embedding for Hyper-Networks (AAAI 2018 · similarity 0.87)
- SNEQ: Semi-Supervised Attributed Network Embedding with Attention-Based Quantisation (AAAI 2020 · similarity 0.87)
- How Low Can You Go? Searching for the Intrinsic Dimensionality of Complex Networks using Metric Node Embeddings (ICLR 2025 · similarity 0.87)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.