Research map / Learning theory and algorithms / Matrix and tensor methods
Binary embeddings with structured hashed projections
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 |
|---|---|---|---|
| Replace the fully random projection with structured projections | MNIST neural network · MNIST test, k=1024 | test error (%): 2.78 7 0.10 -> 3.16 7 0.19 (Toeplitz); 2.78 7 0.10 -> 6.79 7 0.49 (BinCirc) | Mixed |
| Use BinPerm instead of a fully random projection | MNIST neural network · MNIST test, k=1024 | test error (%): 2.78 7 0.10 -> 3.69 7 0.21 | No clear effect |
| Replace the random projection with structured projections | MNIST 1-NN · MNIST test, k=1024 | test error (%): 4.83 7 0.19 -> 6.02 7 0.64 (Circulant); 4.83 7 0.19 -> 12.77 7 2.86 (BinCirc) | Hurt |
| Use the extended hashing pipeline instead of the short pipeline (ablation) | Structured hashed projections · MNIST neural network | test error | No clear effect |
Closest papers by meaning
- The Unreasonable Effectiveness of Structured Random Orthogonal Embeddings (NeurIPS 2017 · similarity 0.87)
- Node Similarities under Random Projections: Limits and Pathological Cases (ICLR 2025 · similarity 0.86)
- Faster Binary Embeddings for Preserving Euclidean Distances (ICLR 2021 · similarity 0.85)
- Efficient end-to-end learning for quantizable representations (ICML 2018 · similarity 0.85)
- Binary Embedding with Additive Homogeneous Kernels (AAAI 2017 · similarity 0.85)
- Learning Discrete Structured Representations by Adversarially Maximizing Mutual Information (ICML 2020 · similarity 0.85)
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.