Unverified50% confidenceFactExact time
Non-stationary time series exhibit a scale drift problem under multi-scale frameworks, that is, the period length or trend slope at different temporal resolutions may change with the environment, causing the information passed from coarse-scale predictions to fine scales to deviate.
1
Sources
50%
Confidence
Long-term
Relevance
11/17/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedLine graphs are sensitive to numerical changes such as amplitude scaling and offset, whereas delay embeddings are insensitive to amplitude and offset, and the two exhibit opposite invariance profiles.74% similarUnverifiedTraditional time series forecasting models face a misalignment between multi-scale inputs and single-scale outputs, which limits forecasting capability.73% similarUnverifiedExisting adaptive-rank methods typically adjust rank based on gradient norms or activation statistics, and these signals are separated from whether the current rank is sufficient by multiple layers of nonlinear transformations, making rank changes hard to trace to their causes.72% similarUnverifiedA shift with a large statistical distance may be absorbed by the equivariance of instance normalization, whereas a shift with a small statistical distance that breaks the equivariance assumption (such as changing the sequence shape or frequency structure) instead causes significant degradation.71% similarUnverifiedFast-motion scenes also add a temporal sampling problem: when inter-frame displacement is too large, the model struggles to establish reliable correspondences between adjacent frames, resulting in ghosting and jitter.70% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1179847API
curl https://kongchang.com/api/v1/knowledge/claims/1179847MCP
get_claim(id=1179847)