Time Series Foundation Models
PulseAugur coverage of Time Series Foundation Models — every cluster mentioning Time Series Foundation Models across labs, papers, and developer communities, ranked by signal.
6 day(s) with sentiment data
-
New OATS strategy enhances Time Series Foundation Models with dynamic data augmentation
Researchers have developed OATS, a novel online data augmentation strategy for Time Series Foundation Models (TSFMs). This method dynamically generates synthetic data tailored to specific training stages, using valuable…
-
New regularization technique combats suboptimal collapse in time series models
Researchers have introduced a new technique called Ground-Truth Neighborhood Regularization (GTN-R) to improve the performance of time series foundation models (TSFMs) when using reinforcement learning (RL) for post-tra…
-
New framework personalizes federated adaptation for time-series models
Researchers have developed a personalized federated sparse adaptation framework for time-series foundation models (TSFMs), aiming to improve energy forecasting by addressing the non-IID nature of private, distributed me…
-
LLM as Forecasting Planner framework integrates LLMs with TSFMs for improved forecasting
Researchers have developed a novel framework called LLM as Forecasting Planner (rc) that integrates large language models (LLMs) with time-series foundation models (TSFMs) for improved text-conditioned forecasting. This…
-
Time series foundation models offer new capabilities for industrial AI
Time series foundation models are emerging as a critical advancement beyond traditional Large Language Models, particularly for industrial AI applications. These models are designed to learn general temporal patterns fr…
-
Time Series Foundation Models show promise for wearable HRV forecasting
A new research paper explores the effectiveness of Time Series Foundation Models (TSFMs) for forecasting heart rate variability (HRV) from consumer wearable devices. The study evaluated TimesFM, Chronos, and MOIRAI agai…
-
New framework unifies post-training methods for Time Series Foundation Models
A new research paper introduces a unifying framework for post-training methods applied to Time Series Foundation Models (TSFMs). The paper categorizes these methods into five types: parameter adaptation, context augment…
-
New frameworks adapt foundation models for drought forecasting · 2 sources tracked
Researchers have developed novel inference-time frameworks, RGMR and SMR^2/MBB, to adapt pre-trained foundation models for regional climate forecasting, specifically for drought prediction. These methods allow for struc…
-
New RMISC Corpus Enhances Time Series Foundation Models with Real-World Data · 2 sources tracked
Researchers have introduced RMISC, a large-scale, real-world corpus designed for training time series foundation models (TSFMs). This corpus, comprising approximately 200 datasets and 142 billion time points, aims to ad…
-
Time Series Foundation Models show promise in electricity price forecasting
A new research paper evaluates the performance of Time Series Foundation Models (TSFMs) in electricity price forecasting, a domain characterized by complex temporal dependencies and distributional shifts. The study intr…
-
Time Series Models Evaluated for US Influenza Forecasting
A new research paper evaluates various time series forecasting models for predicting seasonal influenza in the United States. The study found that a mixture-of-experts model, which combines multiple pretrained forecaste…
-
New framework distills foundation models for specialized time-series forecasting
Researchers have developed a novel framework called Guard to distill knowledge from large, general-purpose foundation models (FMs) into lightweight, specialized time-series forecasters. This approach addresses the chall…
-
Time series model benchmarks may hide critical failures, study finds
A new research paper published on arXiv highlights potential shortcomings in current benchmarks for time series foundation models (TSFMs). The study, focusing on traffic speed forecasting, reveals that aggregate metrics…
-
NYC congestion pricing boosts transit use, study finds
A new study published on arXiv utilizes time series foundation models to analyze the impact of New York City's congestion pricing program, implemented in January 2025. The research found that bus and subway ridership in…
-
New TS-Memory Adapter Enhances Time Series Foundation Models
Researchers have developed TS-Memory, a novel plug-and-play memory adapter designed to enhance Time Series Foundation Models (TSFMs). This method addresses the challenges of adapting TSFMs to new domains by mitigating c…
-
Vision-Language Models Serve as Judges for Time Series Forecasting
Researchers have introduced TimeVista, a new framework that utilizes Vision-Language Models (VLMs) to evaluate time series forecasting. This approach leverages VLMs' ability to interpret time series plots alongside text…
-
New ORCA Method Adapts Time Series Models in Black-Box Settings
Researchers have developed ORCA (Online Residual Contextual Adaptation), a novel method for adapting Time Series Foundation Models (TSFMs) in a black-box setting. This approach focuses on learning from the predictive er…
-
New Agentic Framework Enhances Time Series Forecasting with LLMs
Researchers have introduced KairosAgent, a new framework designed to improve multimodal time series forecasting. This agentic system combines a Large Language Model (LLM) for semantic reasoning with a Time Series Founda…
-
New method audits time series foundation models for data contamination
Researchers have introduced TSFMAudit, a novel method designed to detect data contamination in time series foundation models (TSFMs). This is the first study to address pretraining contamination auditing specifically fo…
-
New AI methods enhance time series forecasting accuracy and interpretability
Researchers have introduced several new methods for time-series forecasting, aiming to improve accuracy and generalization. MeLISA, a latent-free autoregressive model, enhances rollout efficiency and long-horizon statis…