Research map / Label-efficient and robust learning / Recognition and long-tail
Feature Importance Explanations for Temporal Black-Box Models
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 |
|---|---|---|---|
| Use temporal window-aware permutation importance with hypothesis testing and FDR control instead of PERM's individual-timestep permutations | Synthetic classification models with 10 features · Synthetic classification, 10 features | feature power: 0.836 -> 0.930 | Helped |
| Use temporal window-aware permutation importance with hypothesis testing and FDR control instead of PERM's individual-timestep permutations | Synthetic classification models with 10 features · Synthetic classification, 10 features | feature FDR: 0.164 -> 0.037 | Helped |
| Use temporal window-aware permutation importance with hypothesis testing and FDR control instead of PERM's individual-timestep permutations | Synthetic classification models with 10 features · Synthetic classification, 10 features | timestep power: 0.818 -> 0.923 | Helped |
| Use temporal window-aware permutation importance with hypothesis testing and FDR control instead of PERM's individual-timestep permutations | Synthetic classification models with 10 features · Synthetic classification, 10 features | timestep FDR: 0.182 -> 0.054 | Helped |
| Use temporal window-aware permutation importance with hypothesis testing and FDR control instead of PERM's individual-timestep permutations | Synthetic models with 30 features · Synthetic models, 30 features | feature power: 0.830 -> 0.914 | Helped |
| Use temporal window-aware permutation importance with hypothesis testing and FDR control instead of PERM's individual-timestep permutations | Synthetic models with 30 features · Synthetic models, 30 features | feature FDR: 0.170 -> 0.033 | Helped |
| Use temporal window-aware permutation importance with hypothesis testing and FDR control instead of PERM's individual-timestep permutations | Synthetic models with 30 features · Synthetic models, 30 features | timestep power: 0.792 -> 0.909 | Helped |
| Use temporal window-aware permutation importance with hypothesis testing and FDR control instead of PERM's individual-timestep permutations | Synthetic models with 30 features · Synthetic models, 30 features | timestep FDR: 0.208 -> 0.037 | Helped |
| Use TIME-selected features and windows rather than randomly selected features and windows | LSTM · MIMIC-III in-hospital mortality test set | AUROC: 0.801 -> 0.836 | Helped |
| Restrict training to TIME-selected features and windows instead of the full original inputs | LSTM · MIMIC-III in-hospital mortality test set | AUROC: 0.838 -> 0.836 | No clear effect |
Closest papers by meaning
- The Out-of-Distribution Problem in Explainability and Search Methods for Feature Importance Explanations (NeurIPS 2021 · similarity 0.88)
- Temporal Dependencies in Feature Importance for Time Series Prediction (ICLR 2023 · similarity 0.88)
- Testing Semantic Importance via Betting (NeurIPS 2024 · similarity 0.87)
- ReX: A Framework for Incorporating Temporal Information in Model-Agnostic Local Explanation Techniques (AAAI 2025 · similarity 0.85)
- Regional Explanations: Bridging Local and Global Variable Importance (NeurIPS 2025 · similarity 0.85)
- Conditional Feature Importance with Generative Modeling Using Adversarial Random Forests (AAAI 2025 · 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.