Research map / Learning theory and algorithms / Sample complexity
Dimension Estimation Using Random Connection Models
Paulo Serra, Michel Mandjes
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 |
|---|---|---|---|
| increase m_n from log(n) to n (ablation) | canonical estimator · Sierpinski carpet, n=2560 | execution time multiplier: 638.54 times slower | Hurt |
| increase m_n from log(n) to n (ablation) | canonical estimator · Sierpinski carpet, n=2560 | percent of runs with true d within 2 sd: 84.95 -> 0.00 | Hurt |
| increase m_n from log(n) to n (ablation) | canonical estimator · Sierpinski carpet, n=2560 | standard deviation of estimate: 0.22 -> 0.01 | Improved |
| increase m_n (log(n) to n) (ablation) | canonical estimator · Sierpinski carpet, n=2560 | mean estimate (bias): 1.6790 -> 1.6772 | No clear effect |
| add multiplicative bias correction (ablation) | canonical estimator · Gaussian design, d=10, n=10^7 | estimated dimension (true 10): 8.97 -> 9.74 | Improved |
| add exact g_eps(d) additive correction (ablation) | canonical estimator · Gaussian design, d=10, n=10^7 | estimated dimension (true 10): 8.97 -> 9.39 | Improved |
| add two-standard-deviation shift (ablation) | canonical estimator · Gaussian design, d=10, n=10^7 | estimated dimension (true 10): 8.97 -> 10.47 | Improved |
| use exact additive correction instead of multiplicative (ablation) | bias-corrected estimator · Gaussian design | estimated dimension | Hurt |
| increase true dimension d to 10 | canonical estimator · Gaussian design, n=10^7 | estimated dimension: 10 -> 8.97 | Hurt |
| increase sample size n from 10^3 to 10^7 (ablation) | canonical estimator · Gaussian design, d=10 | estimated dimension: 5.95 -> 8.97 | Improved |
| add noise dimensions (lower SNR) | canonical estimator · Gaussian in R^5, n=10^5 | estimated dimension | Mixed |
Closest papers by meaning
- Breaking the curse of dimensionality in structured density estimation (NeurIPS 2024 · similarity 0.84)
- Diffusion Models Encode the Intrinsic Dimension of Data Manifolds (ICML 2024 · similarity 0.84)
- Enhancing Sufficient Dimension Reduction via Hellinger Correlation (ICML 2024 · similarity 0.83)
- A Geometric View of Data Complexity: Efficient Local Intrinsic Dimension Estimation with Diffusion Models (NeurIPS 2024 · similarity 0.83)
- Randomized Dimensionality Reduction for Euclidean Maximization and Diversity Measures (ICML 2025 · similarity 0.83)
- A Wiener Process Perspective on Local Intrinsic Dimension Estimation Methods (AAAI 2025 · similarity 0.83)
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.
