Temporal Fusion Transformers for interpretable multi-horizon time series forecasting
PulseAugur coverage of Temporal Fusion Transformers for interpretable multi-horizon time series forecasting — every cluster mentioning Temporal Fusion Transformers for interpretable multi-horizon time series forecasting across labs, papers, and developer communities, ranked by signal.
3 day(s) with sentiment data
-
Global Temporal Fusion Transformer improves wind power forecasting accuracy
A new study published on arXiv proposes a global Temporal Fusion Transformer (TFT) model for short-term wind power forecasting. This model, trained on a heterogeneous portfolio of wind farms in Germany, aims to outperfo…
-
Federated learning models can learn from static context for time series forecasting
A new research paper explores how federated learning models can effectively utilize static contextual information for time series forecasting. The study found that per-batch aggregation methods like FedAdam are superior…
-
Hybrid SEIR-TFT framework enhances pandemic forecasting with explainable AI
Researchers have developed a hybrid forecasting framework that combines an SEIR epidemiological model with a Temporal Fusion Transformer (TFT) to predict pandemic spread. This approach integrates epidemiological data, i…
-
Hybrid TFT-NBEATS framework enhances solar power forecasting accuracy
Researchers have developed a new hybrid deep learning framework that combines Temporal Fusion Transformers (TFT) and N-BEATS for improved photovoltaic power forecasting. This model effectively captures both short-term f…
-
New framework optimizes 5G resource allocation using probabilistic forecasting
Researchers have developed a new goal-oriented probabilistic forecasting framework to optimize resource allocation in 5G networks. This approach uses the Pinball Loss function with models like DeepAR and Temporal Fusion…
-
Decathlon adopts Time Series Foundation Models for scalable demand forecasting
Decathlon is transitioning its demand forecasting operations to Time Series Foundation Models (TSFMs) to address the scalability and efficiency challenges of managing forecasts for tens of thousands of products across m…
-
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…
-
AI model forecasts airport security throughput using flight schedules
Researchers have developed a novel framework to forecast hourly airport security checkpoint throughput by converting flight schedules into temporally aligned signals. This approach utilizes a Temporal Fusion Transformer…
-
Hierarchical Transformer forecasts emergency department demand coherently
Researchers have developed HierSTT, a novel hierarchical Transformer-based framework designed for coherent forecasting of emergency department (ED) demand across multiple levels. This model jointly predicts hospital, re…
-
New Spatially-Enhanced Transformer Model Improves Prediction for Dynamical Systems
Researchers have developed the Spatially-Enhanced Temporal Fusion Transformer (SE-TFT), an extension of the Temporal Fusion Transformer (TFT) model. This new model is designed to predict multiple outputs for parametric …
-
New AI framework boosts fault prediction accuracy for complex systems
Researchers have developed a novel prognostic framework integrating Spatiotemporal Permutation Entropy (STPE) with Boosted Enhanced Quantile Regression Neural Networks (B-EQRNNs) for long-horizon fault prediction in com…
-
New research tackles uncertainty in smart building load forecasting
A new research paper explores probabilistic load forecasting for smart buildings, focusing on how to handle uncertainty introduced by reconstructed input features. The study compares a post-hoc residual-quantile method …
-
New research tackles time series forecasting with hierarchical and spectral fusion methods · 2 sources tracked
Two new research papers propose novel approaches to time series forecasting. The first, Hierarchical Temporal Fusion (HTF), extends the Temporal Fusion Transformer to ensure coherence in hierarchical data by embedding c…
-
New GeoCat Network Improves IVUS Image Segmentation for Clinical Accuracy
Researchers have developed GeoCat, a novel geometry-consistent network designed for robust segmentation of intravascular ultrasound (IVUS) images. This model addresses limitations in standard methods that often lead to …
-
Deep learning model forecasts Alzheimer's progression with uncertainty estimation · 4 sources tracked
Researchers have developed a deep learning framework to forecast Alzheimer's disease progression with improved accuracy and uncertainty estimation. This probabilistic model, adapted from a Temporal Fusion Transformer, p…
-
AI framework enhances cross-building energy forecasting with transfer learning
Researchers have developed a new transfer learning framework for energy forecasting across different buildings, utilizing the Temporal Fusion Transformer (TFT). This approach aims to improve scalability and robustness f…
-
Deep learning model forecasts climate tipping events with 465x speedup
Researchers have developed a deep learning model, a Temporal Fusion Transformer (TFT), to emulate complex climate simulations. This model can forecast critical climate tipping events, such as ocean collapses, with high …