Research map / Graph learning / Graph algorithms and clustering
DiSC: Differential Spectral Clustering of Features
Ram Dyuthi Sristi, Gal Mishne, Ariel Jaffe
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (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 |
|---|---|---|---|
| Use DiSC instead of Diffusion Maps | MNIST 10-class classification pipeline · MNIST | Test Accuracy (%): 87.42 -> 89.75 | Improved |
| Use DiSC instead of EN-logistic | MNIST 10-class classification pipeline · MNIST | Test Accuracy (%): 82.42 -> 89.75 | Improved |
| Use DiSC instead of Entire Data clustering | MNIST 10-class classification pipeline · MNIST | Test Accuracy (%): 76.23 -> 89.75 | Improved |
| Use DiSC instead of EN | MNIST 10-class classification pipeline · MNIST | Test Accuracy (%): 69.96 -> 89.75 | Improved |
| Set d_A = d_B = 10 rather than 20 (ablation) | DiSC · MNIST digit-pair classification | Test Accuracy: 96.5 % -> 95.2% | Hurt |
| Set d_A = d_B = 30 rather than 20 (ablation) | DiSC · MNIST digit-pair classification | Test Accuracy: 96.5 % -> 96.5% | No clear effect |
| Set d_A = d_B = 40 rather than 20 (ablation) | DiSC · MNIST digit-pair classification | Test Accuracy: 96.5 % -> 96.4% | No clear effect |
| Set d_A = d_B = 50 rather than 20 (ablation) | DiSC · MNIST digit-pair classification | Test Accuracy: 96.5 % -> 94.7 % | Hurt |
| Set d_A = d_B = 60 rather than 20 (ablation) | DiSC · MNIST digit-pair classification | Test Accuracy: 96.5 % -> 88.5% | Hurt |
| Use DiSC rather than DM-changing data under added Gaussian noise | Hyperspectral imaging change detection · Noisy hyperspectral data | Tarp feature grouping robustness | Improved |
| Use differential features rather than diffusion-map features | Feature extraction for MNIST classification · MNIST | Test Accuracy (%): 87.42 -> 89.75 | Improved |
Later papers that compared with this method
3 later accepted papers report DisC as a baseline in their results tables; 3 of them report a better number on most of the results they share with it.
- Disentangling Invariant Subgraph via Variance Contrastive Estimation under Distribution Shifts (ICML 2025 · better on 9 of 9 shared results)
- Subgraph Aggregation for Out-of-Distribution Generalization on Graphs (AAAI 2025 · better on 1 of 1 shared result)
- Does Invariant Graph Learning via Environment Augmentation Learn Invariance? (NeurIPS 2023 · better on 1 of 1 shared result)
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- SpectralNet: Spectral Clustering using Deep Neural Networks (ICLR 2018 · similarity 0.86)
- Robust Spectral Clustering with Rank Statistics (JMLR 2024 · similarity 0.86)
- High-Dimensional Inference for Cluster-Based Graphical Models (JMLR 2020 · similarity 0.85)
- Fast Multi-view Discrete Clustering via Spectral Embedding Fusion (TPAMI 2025 · similarity 0.84)
- Robust Discriminative Clustering with Sparse Regularizers (JMLR 2017 · similarity 0.84)
- A Simple Approach to Automated Spectral Clustering (NeurIPS 2022 · similarity 0.84)
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.
