arXiv：cs.LG· Anidipta Pal·· 3 小时前AI 评分36
SSU-LSF：面向陆地表面预测中非平稳偏差的状态空间遗忘框架
State-Space Unlearning for Non-Stationary Bias in Land Surface Forecasting
AI 导读
研究者提出 SSU-LSF，首个专为基于 Mamba 的地学状态空间模型设计的机器遗忘框架，用于消除陆地表面预测中非平稳混杂事件带来的偏差。
正文
Abstract:Operational land surface forecasting systems built on Mamba-family Structured State Space Models absorb non-stationary confounding events (unrecorded irrigation booms, dam-operation shifts, sensor recalibrations) into their state-transition matrices, silently biasing NDVI, LST, and crop phenology predictions long after the physical cause ends. This paper introduces SSU-LSF (State-Space Unlearning for Land Surface Forecasting), the first machine-unlearning framework purpose-built for geoscientific Mamba-based SSMs. We develop EKFac influence functions specialized to the Mamba state matrices via a closed-form matrix-exponential gradient, use spectral-radius-weighted elbow thresholding to localize a temporal confounding footprint $\Phi$, and apply Hessian-free projected gradient ascent within a KL-divergence trust region augmented by spatial total-variation (TV) regularization. Proposition 1 establishes that residual confounding is bounded by $\mathcal{O}\big((1-\rho(\bar{A})^{T_c})/((1-\rho(\bar{A}))\mu)\big)$, which grows with the window length $T_c$. Across three heterogeneous benchmarks and eleven baselines, SSU-LSF achieves confounding reduction rates of $0.773$ (CropHarvest), $0.821$ (NDVI-LST), and $0.859$ (ERA5), with worst-case clean-domain RMSE degradation of $4.2\%$ on ERA5, converging in 3--5 epochs at $8.4\times$ lower GPU-cost per unlearning request than full retraining. Code: this https URL
| Subjects: | Machine Learning (cs.LG) |
| Cite as: | arXiv:2610.02248 [cs.LG] |
| (or arXiv:2610.02248v1 [cs.LG] for this version) | |
| https://doi.org/10.48550/arXiv.2610.02248 arXiv-issued DOI via DataCite |
Submission history
From: Anidipta Pal [view email]
[v1]
Wed, 30 Sep 2026 18:27:59 UTC (1,481 KB)
来源：arXiv：cs.LG · arxiv.org