Chronos-2 Forecasting Model
PulseAugur coverage of Chronos-2 Forecasting Model — every cluster mentioning Chronos-2 Forecasting Model across labs, papers, and developer communities, ranked by signal.
6 day(s) with sentiment data
-
Advanced AI time-series models fail to beat simple baselines in market prediction
A recent benchmark study found that prominent time-series foundation models, including TimesFM 3.0, Chronos-2, and Toto 2.0, failed to outperform simple baseline strategies when applied to market time series data. The e…
-
AI models improve sensor data forecasting and fault detection · 2 sources tracked
Researchers have developed methods to improve the accuracy and robustness of time-series forecasting models, particularly for sensor data. One study demonstrates that Long-Short Term Memory (LSTM) network ensembles can …
-
Time-Series Foundation Models Tested for Sampling Frequency Invariance
A new study evaluates the sampling frequency invariance of four prominent Time-Series Foundation Models (TSFMs): MOMENT, Chronos-2 Forecasting Model, Moirai-MoE, and TimesFM. The research introduces novel metrics to qua…
-
Chronos-2 foundation model advances multivariate time series forecasting
Researchers have introduced Chronos-2, a new foundation model designed for multivariate time series forecasting. This model is capable of modeling interactions between multiple time series and external factors, enabling…
-
New FARE method boosts retail forecasting accuracy by 10-14%
Researchers have developed FARE (Factor-Aware Retrieval with Error-Guided Fusion), a novel method for improving retail sales forecasting. FARE addresses the challenge of incomplete factor combinations in historical data…
-
TSFM carbon footprint analyzed; pruning offers emissions reduction
A new research paper explores the sustainability of Time Series Foundation Models (TSFMs), specifically analyzing their carbon footprint during pretraining and deployment. The study found that while TSFMs like Chronos-2…
-
Time series models answer associations, not interventions, study finds
A new study published on arXiv examines the capabilities of covariate-aware time series foundation models (TSFMs) in answering "what-if" questions. The research found that models like Chronos-2, TiRex-2, TabPFN-TS, and …
-
New dynamic decoding strategy boosts time-series model forecasting accuracy
Researchers have developed a novel dynamic multi-step decoding strategy for time-series foundation models, aiming to improve forecasting accuracy. This approach adaptively selects the optimal decoding method for each fo…
-
New research explores efficient adaptation of time series foundation models for electrical load forecasting
A new research paper explores parameter-efficient adaptation techniques for time series foundation models (TSFMs) in electrical load forecasting. The study compares pretrained inference, full fine-tuning, and low-rank a…
-
Chronos-2 and TimesNet benchmarked for load forecasting under uncertainty
A new paper benchmarks time series foundation models (TSFMs) for short-term load forecasting (STLF) under varying levels of covariate uncertainty. The study found that Chronos-2 performed best when future covariates wer…
-
New research tackles in-context learning for time-series AI models
Researchers are exploring methods to improve the in-context learning capabilities of time-series foundation models (TSFMs). A new paper investigates how models like Chronos-2, TimesFM-2.5, and TabPFN-TS handle counterfa…
-
Pakistan oil firm designs sovereign AI for HSE incident prediction
An oil and gas operator in Pakistan requires an air-gapped AI system for its health, safety, and environment (HSE) department to predict incidents before they occur. The proposed architecture emphasizes self-hosted mode…
-
Time Series FMs show surprising blind spots in basic temporal reasoning
A new research paper titled "Foundations without Fundamentals: Zero-Shot Blind Spots in Time Series FMs" highlights significant limitations in current Time Series Foundation Models (TSFMs). Despite their success on broa…
-
Group Attention Mechanism Analysis Reveals Mixed Performance in Forecasting
Researchers have analyzed the group attention mechanism, originally developed for the Chronos-2 Forecasting Model, to understand its effectiveness in multivariate and in-context learning scenarios. Their findings indica…
-
Chronos-2 model excels in peak-aware electricity load forecasting
A new research paper introduces the Peak-Aware Short-Term Load Forecasting (STLF) framework, designed to improve accuracy during high-demand periods for distribution grid operators. The study compares various models, in…
-
Foundation models show mixed results for pedestrian crowd forecasting
A new study published on arXiv evaluates the effectiveness of time-series foundation models (FMs) for pedestrian crowd count forecasting. The research compares seven different forecasting approaches, including tradition…
-
New research tackles intermittent demand forecasting challenges · 2 sources tracked
Two new research papers explore the challenges of intermittent demand forecasting, where demand occurs infrequently and time series often contain many zero observations. The first paper, "Accuracy Is Not Service," intro…
-
FlowTSFM introduces novel recurrent transport for time series models
Researchers have introduced FlowTSFM, a novel encoder architecture for time series foundation models that utilizes depth as a recurrent transport process. Instead of multiple independent Transformer layers, FlowTSFM emp…
-
New ensemble method improves photovoltaic power forecasting accuracy
Researchers have developed a novel hierarchical ensemble method for short-term photovoltaic power forecasting. This approach combines various models, including temporal neural networks, historical analogs, climatology, …
-
AWS integrates Databricks and Amazon Quick for automated retail replenishment
AWS has detailed a new system for automating retail replenishment by integrating Databricks and Amazon Quick. This system uses Databricks' Many Model Forecasting (MMF) with the Chronos-2 model to predict demand for SKUs…