LeapTS reformulates forecasting as adaptive multi-horizon scheduling via hierarchical control and NCDEs, delivering at least 7.4% better performance and 2.6-5.3x faster inference than Transformer baselines while adapting to non-stationary dynamics.
Autoformer: Decomposition transformers with auto-correlation for long-term series forecasting
7 Pith papers cite this work. Polarity classification is still indexing.
7
Pith papers citing it
representative citing papers
Prime attention modulates attention keys and values per channel-pair and reports improved MTS forecasting accuracy across several benchmarks.
A seq2seq model with a new adaptive-frequency layer learns the inverse mapping from flapping-wing forces to wing kinematics, improving median prediction error over transformer baselines by about 11% on two datasets.
Retrieving similar time-series segments from a multi-domain knowledge base and injecting them through a learned Channel Prompting module improves zero-shot forecasting of a frozen TSFM, though gains are small and leakage risks are unaddressed.
The paper proposes causal-Mamba, a joint treatment-outcome deep sequential model, to estimate average treatment effects of Google Trends-like external signals on social media engagement and to rank influential users.
A simple gradient-free 'directional gradient approximation' attack makes LLM time series forecasters degrade more than equivalent random noise, across GPT-3.5, GPT-4, LLaMa, Mistral, TimeGPT, and TimeLLM.
citing papers explorer
-
LeapTS: Rethinking Time Series Forecasting as Adaptive Multi-Horizon Scheduling
LeapTS reformulates forecasting as adaptive multi-horizon scheduling via hierarchical control and NCDEs, delivering at least 7.4% better performance and 2.6-5.3x faster inference than Transformer baselines while adapting to non-stationary dynamics.
-
Dynamic Relational Priming Improves Transformer in Multivariate Time Series
Prime attention modulates attention keys and values per channel-pair and reports improved MTS forecasting accuracy across several benchmarks.
-
A Deep Inverse-Mapping Model for a Flapping Robotic Wing
A seq2seq model with a new adaptive-frequency layer learns the inverse mapping from flapping-wing forces to wing kinematics, improving median prediction error over transformer baselines by about 11% on two datasets.
-
TimeRAF: Retrieval-Augmented Foundation model for Zero-shot Time Series Forecasting
Retrieving similar time-series segments from a multi-domain knowledge base and injecting them through a learned Channel Prompting module improves zero-shot forecasting of a frozen TSFM, though gains are small and leakage risks are unaddressed.
-
Estimating Online Influence Needs Causal Modeling! Counterfactual Analysis of Social Media Engagement
The paper proposes causal-Mamba, a joint treatment-outcome deep sequential model, to estimate average treatment effects of Google Trends-like external signals on social media engagement and to rank influential users.
-
Adversarial Vulnerabilities in Large Language Models for Time Series Forecasting
A simple gradient-free 'directional gradient approximation' attack makes LLM time series forecasters degrade more than equivalent random noise, across GPT-3.5, GPT-4, LLaMa, Mistral, TimeGPT, and TimeLLM.
- What Causes Performance Degradation in Cross-Subject EEG Classification?