AAAI 2025 accepted papers on Time series

Foundation models for time series · 27 papers

• TimeCMA: Towards LLM-Empowered Multivariate Time Series Forecasting via Cross-Modality Alignment (96 citations)
• CALF: Aligning LLMs for Time Series Forecasting via Cross-modal Fine-Tuning (66 citations)
• ChatTime: A Unified Multimodal Time Series Foundation Model Bridging Numerical and Textual Data (28 citations)
• TimeCAP: Learning to Contextualize, Augment, and Predict Time Series Events with Large Language Model Agents (19 citations)
• Geolocation Representation from Large Language Models Are Generic Enhancers for Spatio-Temporal Learning (16 citations)
• Federated Foundation Models on Heterogeneous Time Series (12 citations)
• TimePFN: Effective Multivariate Time Series Forecasting with Synthetic Data (11 citations)
• Battling the Non-stationarity in Time Series Forecasting via Test-time Adaptation (9 citations)
• Hierarchical Classification Auxiliary Network for Time Series Forecasting (7 citations)
• Content-aware Balanced Spectrum Encoding in Masked Modeling for Time Series Classification (6 citations)
• TimeDP: Learning to Generate Multi-Domain Time Series with Domain Prompts (6 citations)
• Frequency-Masked Embedding Inference: A Non-Contrastive Approach for Time Series Representation Learning (5 citations)

and 15 more in the interactive map.

Long-term forecasting · 27 papers

• WPMixer: Efficient Multi-Resolution Mixing for Long-Term Time Series Forecasting (77 citations)
• Unlocking the Power of LSTM for Long Term Time Series Forecasting (71 citations)
• Adaptive Multi-Scale Decomposition Framework for Time Series Forecasting (58 citations)
• Unlocking the Power of Patch: Patch-Based MLP for Long-Term Time Series Forecasting (41 citations)
• Amplifier: Bringing Attention to Neglected Low-Energy Components in Time Series Forecasting (26 citations)
• Affirm: Interactive Mamba with Adaptive Fourier Filters for Long-term Time Series Forecasting (24 citations)
• FilterTS: Comprehensive Frequency Filtering for Multivariate Time Series Forecasting (22 citations)
• MPTSNet: Integrating Multiscale Periodic Local Patterns and Global Dependencies for Multivariate Time Series Classification (20 citations)
• RhythmMamba: Fast, Lightweight, and Accurate Remote Physiological Measurement (18 citations)
• Times2D: Multi-Period Decomposition and Derivative Mapping for General Time Series Forecasting (11 citations)
• Revisiting Attention for Multivariate Time Series Forecasting (9 citations)
• TimeCHEAT: A Channel Harmony Strategy for Irregularly Sampled Multivariate Time Series Analysis (9 citations)

and 15 more in the interactive map.

Causal and generative time series · 16 papers

• Auto-Regressive Moving Diffusion Models for Time Series Forecasting (22 citations)
• Self-attention-based Diffusion Model for Time-series Imputation in Partial Blackout Scenarios (9 citations)
• Cherry-Picking in Time Series Forecasting: How to Select Datasets to Make Your Model Shine (7 citations)
• DualDynamics: Synergizing Implicit and Explicit Methods for Robust Irregular Time Series Analysis (7 citations)
• Neural Conformal Control for Time Series Forecasting (5 citations)
• Population Aware Diffusion for Time Series Generation (5 citations)
• DroughtSet: Understanding Drought Through Spatial-Temporal Learning (4 citations)
• Kernel Representation Learning for Time Sequence: Algorithm, Theory, and Applications (3 citations)
• ERCI: An Explainable Experience Replay Approach with Causal Inference for Deep Reinforcement Learning (2 citations)
• Probabilistic Forecasting of Irregularly Sampled Time Series with Missing Values via Conditional Normalizing Flows (2 citations)
• Mixture of Online and Offline Experts for Non-Stationary Time Series (1 citation)
• Modeling Latent Non-Linear Dynamical System over Time Series (1 citation)

and 4 more in the interactive map.

Anomaly detection · 3 papers

• Graph Mixture of Experts and Memory-augmented Routers for Multivariate Time Series Anomaly Detection (22 citations)
• GCAD: Anomaly Detection in Multivariate Time Series from the Perspective of Granger Causality (18 citations)
• TAIL-MIL: Time-Aware and Instance-Learnable Multiple Instance Learning for Multivariate Time Series Anomaly Detection (5 citations)