Results 211 to 220 of about 1,858,266 (298)

Optimization of hyperparameters of deep learning models

open access: yes
Les modèles d'apprentissage profond ont permis de grandes percées dans de nombreux domaines tels que le traitement de texte, d'image et du son. Ils se distinguent des autres modèles grâce à leurs grandes capacités de représentation automatique.Ces ...
Lacharme, Guillaume
core  

Fusing Direct and Indirect Measurements Through Multi‐Fidelity Learning For Accelerated Electrocaloric Materials Discovery

open access: yesAdvanced Science, EarlyView.
A multi‐fidelity framework integrates sparse direct and abundant indirect electrocaloric measurements. Multi‐objective active learning accelerates BaTiO3‐based electrocaloric materials discovery at –70∘C$^{\circ }{\rm C}$. A diffuse transition enables an electrocaloric strength of 0.06×$\times$10−6 Km/V at –70℃ with an operational temperature span of ...
Bo Wang   +8 more
wiley   +1 more source

An Interpretable, Data‐Driven, Hierarchical Multi‐Domain Fusion Framework for Classification and Motor Function Scoring in Chronic Ankle Instability

open access: yesAdvanced Science, EarlyView.
An AI‐enabled digital twin framework integrates wearable EMG sensing with hierarchical multi‐domain fusion to classify chronic ankle instability, distinguish clinically relevant subtypes, and generate continuous motor function scores. Clinically interpretable functional stratification and SHAP‐based biomarker analysis provide transparent decision ...
Tianle Jie   +12 more
wiley   +1 more source

A Bioinspired, Multimodal Soft Tactile Skin with Task‐Adaptive Perception for Intelligent Robotic Manipulation

open access: yesAdvanced Science, EarlyView.
Perception in robotic manipulation improves through the cooperative and complementary operation of biologically inspired multimodal tactile sensors. Receptor‐specific contribution analysis identifies which sensing modalities are required for specific tactile tasks.
Yu‐Jin Lee   +4 more
wiley   +1 more source

A Site‐Aware Representation Learning Framework For Unified Molecular Interaction Modeling and Generative Design

open access: yesAdvanced Science, EarlyView.
MolDBG is a site‐aware, sequence‐only framework that unifies drug‐target affinity prediction, binding‐site identification, and affinity‐conditioned molecular generation for structured proteins. Guided by multi‐task binding‐site supervision, it aligns interaction‐critical residues before learning drug‐target representations and simultaneously infers ...
Gang Luo   +6 more
wiley   +1 more source

STWave: Fine‐Scale Spatial Structure Discovery in Microscopic‐Resolution Spatial Transcriptomics via Patchwise Wavelet Graphs

open access: yesAdvanced Science, EarlyView.
STWave transforms massive microscopic‐resolution spatial transcriptomics into interpretable fine‐scale tissue maps through patch‐wise inference, wavelet‐based multi‐scale encoding, and dual‐domain reconstruction. It reduces noise while preserving weak spatial signals, enabling efficient analysis of 6 40 000 spots of 2.47 GB GPU memory and revealing ...
Tao Jiang   +9 more
wiley   +1 more source

Home - About - Disclaimer - Privacy