Koopman
PulseAugur coverage of Koopman — every cluster mentioning Koopman across labs, papers, and developer communities, ranked by signal.
2 day(s) with sentiment data
-
New STEK framework enhances retrospective battery characterization
Researchers have developed STEK, a novel framework designed to improve retrospective battery characterization across diverse cell types. This method integrates a cell-level lifecycle terrain with an endpoint-derived str…
-
New SALTO framework enhances nonlinear event inference using Koopman theory
Researchers have developed SALTO, a novel framework based on Koopman theory for inferring finite-amplitude sources in nonlinear systems. This method improves upon traditional adjoint techniques by enabling the predictio…
-
New kernel method learns Koopman embeddings, rivals ANNs
Researchers have developed a new kernel-based method for learning finite-dimensional Koopman embeddings directly from data, called extended dynamic mode decomposition with kernel-based dictionary learning (EDMD-kDL). Th…
-
KAIROS framework uses Koopman operator for advanced dynamic graph learning
Researchers have developed KAIROS, a novel self-supervised framework for dynamic graph contrastive learning. This method explicitly models temporal evolution by embedding a differentiable Koopman operator, which lineari…
-
New AdaSpecK framework enhances temporal domain generalization for streaming data
Researchers have introduced AdaSpecK, a novel framework designed to improve Temporal Domain Generalization (TDG) for streaming data that experiences distribution shifts over time. This spectral-Koopman framework incorpo…
-
New SADFED framework identifies Koopman models without predefined architectures
Researchers have developed a new framework called Spatially Aware Dictionary-Free Koopman Eigenfunction Identification (SADFED) for discovering Koopman models from data. This method does not require pre-defining a lifti…
-
New methods enhance LLM safety with dynamical systems and low-latency guardrails · 4 sources tracked
Researchers have developed two novel approaches to enhance the safety of Large Language Models (LLMs). The first method, detailed in an arXiv paper, utilizes a dynamical systems framework based on Koopman operators to c…
-
New operator-theoretic bounds for multitask deep learning
Researchers have developed operator-theoretic generalization bounds for deep multitask learning models. The approach represents network layers as Koopman composition operators within vector-valued reproducing kernel Hil…
-
New framework models nonlinear delay dynamics using Koopman operator
Researchers have developed a novel framework for approximating the Koopman operator of nonlinear delay differential equations (DDEs). This approach bridges the gap between infinite-dimensional DDE dynamics and finite-di…
-
New method enhances equation discovery from noisy data using Koopman dynamics
Researchers have developed a dynamics-aware method for identifying governing equations from sparse and noisy data, building upon techniques like Sparse Identification of Nonlinear Dynamics (SINDy) and PDE Functional Ide…
-
New SKooP method boosts reinforcement learning for robot locomotion
Researchers have developed SKooP (Symmetric Koopman Predictions), a novel approach to enhance reinforcement learning for legged robot locomotion. This method combines morphological symmetries with a Koopman model learne…
-
Federated Koopman Learning Framework Enhances IoT Anomaly Detection
Researchers have developed FedKAD, a novel federated learning framework designed for anomaly detection in Internet of Things (IoT) systems. This approach utilizes lightweight Koopman representations to learn normal temp…
-
New method enhances accuracy of Koopman spectral approximations
Researchers have developed a new method for learning neural network dictionaries to improve the accuracy of Koopman spectral approximations in nonlinear dynamics. This approach focuses on minimizing residual errors, whi…
-
New KFTD Network boosts ocean forecasting accuracy and speed
Researchers have developed a new time-continuous forecasting model called the Koopman-Fourier Time-Differentiable (KFTD) Network. This model aims to improve the accuracy and efficiency of ocean spatiotemporal forecastin…
-
Quantum-Informed ML Shows Practical Advantage in Chaos Prediction
Researchers have developed a new theoretical framework for achieving practical quantum advantage in quantum-informed machine learning, specifically for predicting chaotic systems. This approach utilizes higher-order qua…
-
New Method Uses Personalized PageRank to Find Koopman Invariant Subspaces
Researchers have developed a novel method for identifying Koopman invariant subspaces using Personalized PageRank (PPR) applied to Extended Dynamic Mode Decomposition (EDMD) matrices. This technique exploits zero-block …
-
New Mamba model variant enhances memory retention and bilinear computation
Researchers have introduced Bilinear Input Modulation (BIM) to enhance Selective State Space Models (SSMs), specifically Mamba, by incorporating state-input products. This augmentation allows for improved memory retenti…