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
-
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…
-
New Vision-Based Model TITAnD Enables Multi-Month Trajectory Anomaly Detection
Researchers have introduced TITAnD, a novel approach to trajectory anomaly detection that reframes the problem as a computer vision task. By representing trajectories as Hyperspectral Trajectory Images (HTIs), TITAnD un…
-
Amazon's Chronos-2 Model Assessed for Grid Load Forecasting
A new research paper evaluates Amazon's Chronos-2 Forecasting Model for its effectiveness in real-world grid load forecasting. The study found that while Chronos-2 shows promise, especially with task-specific fine-tunin…
-
Decathlon scales demand forecasting with Chronos-2 on AWS
Decathlon, a global sporting goods retailer, has adopted the Chronos-2 forecasting model to improve its demand prediction capabilities. The company evaluated multiple time series foundation models and selected Chronos-2…
-
New framework identifies biases in time series foundation models
A new research paper proposes a causal analysis framework to identify biases and failure modes in time series foundation models before deployment. The study applied this framework to Chronos-2 and TimesFM-2.5, revealing…
-
TabPFN-TS model shows promise for district heating forecasts
A new study systematically evaluates TabPFN-TS, a model utilizing synthetic pretraining data, for probabilistic heat load forecasting in district heating networks. The research compares TabPFN-TS against existing time-s…
-
New LoRA Framework Adapts Chronos-2 for Electricity Price Forecasting
Researchers have developed a novel framework to adapt the Chronos-2 time-series foundation model for day-ahead electricity price forecasting, particularly in markets with limited historical data. This approach utilizes …
-
Migas 1.5 integrates text context into time-series forecasting
Synthefy's Migas 1.5, released in April 2026, introduces a novel feature that allows time-series forecasting models to incorporate plain-text context alongside numerical data. This innovation aims to bridge the gap betw…
-
EU-AI Act compliant forecasting pipeline beats large models
A recent study evaluated a short-term load forecasting pipeline designed to comply with the EU-AI Act's requirements for safety-critical environments. The pipeline, built on the open-source Python library spotforecast2-…
-
Transformer models achieve 10.7% lower error in electrical load forecasting
Researchers have developed a new benchmark for electrical load forecasting across various grid levels, from control areas to individual consumers. Their study found that Transformer-based models, particularly the standa…
-
New framework audits conditional quantile forecasters for miscalibration
Researchers have developed a new framework for continuously auditing conditional quantile forecasters, which are crucial for sequential decisions in areas like supply chain management. This method addresses limitations …
-
Google Research releases TimesFM 2.5 for zero-shot time-series forecasting
Google Research has released TimesFM 2.5, an open-source foundation model for time-series forecasting. This model, with 200 million parameters and a context window of up to 16,384 points, can predict future trends witho…
-
Foundation models struggle with extreme wildfire smoke prediction, study finds
A new study evaluated the generalizability of foundation models for predicting extreme PM2.5 concentrations from wildfire smoke, a critical public health challenge. Researchers compared six time series foundation model …
-
Time series foundation models show promise for energy load forecasting
A new research paper evaluates time series foundation models for low-voltage peak load forecasting in energy systems. The study compares Chronos-Bolt, Chronos-2, and TabPFN-TS against baseline models, finding Chronos-2 …
-
LLMs drive automated feature engineering for structured data
Researchers have developed Evolutionary Feature Engineering (EFE), a novel framework that leverages large language models (LLMs) to automatically discover preprocessing transformations for structured data. EFE represent…