Results 121 to 130 of about 6,478,030 (244)

Improving the Accuracy of Neural Network Pattern Recognition by Fractional Gradient Descent

IEEE Access
In this paper we propose the fractional gradient descent for increasing the training and work of modern neural networks. This optimizer searches the global minimum of the loss function considering the fractional gradient directions achieved by Riemann ...
Ruslan I. Abdulkadirov, Pavel A. Lyakhov, Valentina A. Baboshina, Nikolay N. Nagornov   +3 more
doaj   +1 more source

Approximation and gradient descent training with neural networks

Sampling Theory, Signal Processing, and Data Analysis
It is well understood that neural networks with carefully hand-picked weights provide powerful function approximation and that they can be successfully trained in over-parametrized regimes. Since over-parametrization ensures zero training error, these two theories are not immediately compatible.
openaire   +2 more sources

Neural Networks and the Natural Gradient

, 2010
Neural network training algorithms have always suffered from the problem of local minima. The advent of natural gradient algorithms promised to overcome this shortcoming by finding better local minima. However, they require additional training parameters and computational overhead.
openaire   +2 more sources

Fiber‐Based Wearable Electronics as Adaptive Biointerfaces for Healthcare Applications

Advanced Functional Materials, EarlyView.
Fiber‐based wearable electronics have been extensively studied as adaptive healthcare biointerfaces. In this review, we classify and introduce fiber‐based systems according to biointerface design, conductive materials, transduction mechanisms, and healthcare applications, and provide a perspective on their translation toward reliable, continuous ...
Sangwoo Hong, Jinsol Kim, Yoonjae Jeong, Sejong Kang, Daeho Lee, Seung Hwan Ko   +5 more
wiley   +1 more source

Role of Structural Disorder on Phonon Transport and Thermal Conductivity in Li6PS5Br

Advanced Functional Materials, EarlyView.
Solid electrolytes are fast ion and slow heat conductors. Although both transport properties are governed by lattice dynamics and atomic structure, their interplay remains poorly understood. Here, we study this interrelation in anion‐ordered and disordered Li6PS5Br.
Lukas Ketter, Ryoma Sasaki, Kyra Strotmann, Wolfgang G. Zeier, Yoshitaka Tateyama   +4 more
wiley   +1 more source

Nitrospinics: Nitride Materials for Spin‐Orbit and Orbital Torques From Memory to Artificial Intelligence

Advanced Functional Materials, EarlyView.
Further advances in applications can be anticipated through the development of unique and functional materials beyond conventional alloys. Nitrogen enables tailor‐made orbital textures in various transition‐metal hosts, paving the way for devices ranging from spintronics memory to artificial intelligence hardware.
Prabhat Kumar, Gaurav Kumar Shukla, Yoshio Miura, Shinji Isogami   +3 more
wiley   +1 more source

GeS as an Emerging OTS Material: Atomic Structure and Modeling of ON/OFF Switching

Advanced Functional Materials, EarlyView.
The structural puzzle of glassy GeS, an advanced memory‐selector material, has finally been disentangled, revealing a patchwork of mixed tetrahedral and trigonal species. Under a high electric field, over‐ and under‐coordinated Ge and S defects, together with three‐membered rings, form an interconnected network associated with localized and extended ...
Andrey Tverjanovich, Jörg C. Neuefeind, Chris J. Benmore, Daniele Fontanari, Anton Sokolov, Maxim Khomenko, Shinji Kohara, Aleksei Bytchkov, Takeshi Usuki, Maria Bokova, Mohammad Kassem, Pascal Masselin, Mariana Milochova, Eugene Bychkov   +13 more
wiley   +1 more source

Gradient Inversion Attack on Graph Neural Networks

CoRR
Graph federated learning is of essential importance for training over large graph datasets while protecting data privacy, where each client stores a subset of local graph data, while the server collects the local gradients and broadcasts only the aggregated gradients. Recent studies reveal that a malicious attacker can steal private image data from the
Divya Anand Sinha, Yezi Liu, Ruijie Du, Athina Markopoulou, Yanning Shen   +4 more
openaire   +3 more sources

Neuromorphic Devices for Extreme Environments: State‐Variable Reliability and Stabilization Strategies

Advanced Functional Materials, EarlyView.
Neuromorphic devices for extreme environments are interpreted through the physical states that store and process information. Radiation, temperature extremes, and vacuum can perturb vacancies, filaments, trapped charges, polarization, interfaces, and readout signals.
Taehyun Park, Geon Park, Junhyung Cho, Jaechan Song, Wangmyung Choi, Jaehyun Hur, Nicola Gasparini, Hocheon Yoo   +7 more
wiley   +1 more source

A Model for Programmability and Virtuality in Dynamical Neural Networks [PDF]

, 2009
In this dissertation a fixed-weight architecture for Continuous Time Recurrent Neural Networks (CTRNNs) is proposed in order to give an account for biological phenomena, controlled by neuronal activity, in which changes of behavior occur so fast that ...
Donnarumma, Francesco
core   +1 more source