Research map / Learning theory and algorithms / Matrix and tensor methods
A Random Matrix Approach to Echo-State Neural Networks
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 single Haar reservoir with multi-memory block-diagonal Haar W | linear ESN reservoir · Mackey-Glass one-step prediction | testing NMSE | Helped |
| Use non-normal (i.i.d.) reservoir instead of symmetric Wigner reservoir (ablation) | linear ESN reservoir · Mackey-Glass delay-τ memorization task | training and testing NMSE | Helped |
| Add small internal noise with optimal η² (trade-off with regularization effect) (ablation) | linear ESN with internal noise · Mackey-Glass one-step prediction with 1% or 10% impulsive noise in test inputs | testing NMSE | Helped |
| Choose input vector m as a leading eigenvector of W | linear ESN · theory (referenced to extended version) | testing MSE | Hurt |
| Reduce η² below ~n^(-1/2) (theory approximation regime) | linear ESN theory accuracy · Mackey-Glass one-step prediction | accuracy of deterministic approximation | Hurt |
Closest papers by meaning
- Statistical Mixing Guarantees for Contractive Echo State Networks (NeurIPS 2026 · similarity 0.86)
- Unfolding recurrence by Green’s functions for optimized reservoir computing (NeurIPS 2020 · similarity 0.85)
- Improved memory in recurrent neural networks with sequential non-normal dynamics (ICLR 2020 · similarity 0.85)
- Eigenvalue Normalized Recurrent Neural Networks for Short Term Memory (AAAI 2020 · similarity 0.85)
- Frequency Domain Reservoir Computing (NeurIPS 2026 · similarity 0.84)
- Recurrent neural networks: vanishing and exploding gradients are not the end of the story (NeurIPS 2024 · 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.