Research map / Time series / Foundation models for time series
Interpretable Classification of Categorical Time Series Using the Spectral Envelope and Optimal Scalings
Zeda Li, Scott A. Bruce, Tian Cai
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 |
|---|---|---|---|
| add optimal scalings to the spectral envelope (ablation) | spectral envelope classifier · Simulation Case 3 (N_J=50, T=100) | correct classification rate (%): 71.08 -> 83.79 | Improved |
| add optimal scalings to the spectral envelope (ablation) | spectral envelope classifier · Simulation Case 3 (N_J=50, T=500) | correct classification rate (%): 94.32 -> 98.42 | Improved |
| add spectral envelope to the optimal scalings (ablation) | scalings classifier · Simulation Case 3 (N_J=50, T=100) | correct classification rate (%): 72.91 -> 83.79 | Improved |
| add spectral envelope to the optimal scalings (ablation) | scalings classifier · Simulation Case 1 (N_J=50, T=100) | correct classification rate (%): 49.92 -> 92.99 | Improved |
| add optimal scalings to the spectral envelope (ablation) | spectral envelope classifier · Simulation Case 1 (N_J=50, T=100) | correct classification rate (%): 93.54 -> 92.99 | No clear effect |
| add optimal scalings to the spectral envelope (ablation) | spectral envelope classifier · Simulation Case 2 (N_J=50, T=100) | correct classification rate (%): 50.71 -> 76.01 | Improved |
| add spectral envelope to the optimal scalings (ablation) | scalings classifier · Simulation Case 2 (N_J=50, T=500) | correct classification rate (%): 99.40 -> 94.20 | No clear effect |
| use EnvSca instead of sequence learner | sequence learner · Simulation Case 1 (N_J=50, T=100) | correct classification rate (%): 90.40 -> 92.99 | Improved |
| use EnvSca instead of sequence learner | sequence learner · Simulation Case 2 (N_J=50, T=100) | correct classification rate (%): 77.94 -> 76.01 | Hurt |
| increase series length T from 100 to 500 | EnvSca · Simulation Case 1 (N_J=50) | correct classification rate (%): 92.99 -> 99.56 | Improved |
| increase series length T from 100 to 500 | EnvSca · Simulation Case 2 (N_J=50) | correct classification rate (%): 76.01 -> 94.20 | Improved |
| increase training series per group N_J from 20 to 100 | EnvSca · Simulation Case 3 (T=100) | correct classification rate (%): 81.02 -> 84.97 | Improved |
Closest papers by meaning
- MPTSNet: Integrating Multiscale Periodic Local Patterns and Global Dependencies for Multivariate Time Series Classification (AAAI 2025 · similarity 0.84)
- Frequency-Masked Embedding Inference: A Non-Contrastive Approach for Time Series Representation Learning (AAAI 2025 · similarity 0.83)
- An Unsupervised Approach for Periodic Source Detection in Time Series (ICML 2024 · similarity 0.83)
- Peri-midFormer: Periodic Pyramid Transformer for Time Series Analysis (NeurIPS 2024 · similarity 0.82)
- Diffusion Language-Shapelets for Semi-supervised Time-Series Classification (AAAI 2024 · similarity 0.82)
- Content-aware Balanced Spectrum Encoding in Masked Modeling for Time Series Classification (AAAI 2025 · similarity 0.82)
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.
