Results 101 to 110 of about 25,600 (268)

Identifying Dynamical Quantum Phase Transitions With a Migratable Quantum‐Classical Hybrid Neural Network

Advanced Intelligent Systems, EarlyView.
A hybrid quantum‐classical architecture is introduced to accurately identify dynamical quantum phase transitions from time‐evolved quantum states. The QCNN serves as a quantum dynamical feature extractor, while the classical network learns temporal correlations from a low‐dimensional readout sequence. The framework attains high accuracy, remains robust
Daili Li, Xin–Xin Yang, Xiaoting Wang, Wei Zhang   +3 more
wiley   +1 more source

Progressive Training of Hybrid Clipped Rectified Linear Unit for Resilient Convolution Neural Networks

IEEE Access
Convolutional neural networks used in safety-critical systems must remain reliable under hardware-induced faults. Activation clipping is an effective fault-mitigation technique, yet existing approaches face a fundamental trade-off: layer-wise clipping ...
Seyedhamidreza Mousavi, Mohammad Hasan Ahmadilivani, Jaan Raik, Maksim Jenihhin, Masoud Daneshtalab   +4 more
doaj   +1 more source

The Cost of Avoiding Backpropagation

CoRR
Forward-mode automatic differentiation (FmAD) and zero-order (ZO) optimization have been proposed as memory-efficient alternatives to backpropagation (BP) for gradient computation, especially in low-resource settings. However, their practical benefits remain unclear due to two key gaps: a lack of comparison against memory-efficient BP variants, such as
Kunjal Panchal, Sunav Choudhary, Yuriy Brun, Hui Guan 0001   +3 more
openaire   +2 more sources

Gate‐Align‐SED: Semi‐Supervised Sound Event Detection via Adaptive Feature Gating and Cross‐Task Alignment in Situation Awareness

Advanced Intelligent Systems, EarlyView.
Overview of the proposed Gate‐Align‐SED, including two stages of training: (1) Mean‐Teacher SSL Training; and (2) Enhancer Model Training. In complex real‐world environments such as disaster monitoring, effective sound event detection (SED) is often hindered by the presence of noise and limited labeled data.
Jieli Chen, Li‐Minn Ang, Chee Shen Lim, Kah Phooi Seng, Jeremy Smith   +4 more
wiley   +1 more source

Analysis and Prediction of Vertical Office Network Bandwidth Using a Backpropagation-Based Neural Network

Ilkom Jurnal Ilmiah
The digitalization of business and operational processes in vertical offices has transformed work behaviour, creating a critical need for stable internet connectivity to ensure smooth operations.
Muhammad Jamil Asshiddiq, Iffatul Mardhiyah   +1 more
doaj   +1 more source

Shapley Additive Explanation for Local Class Differentiation: Local Explainability for Class Differentiation in Classification Models

Advanced Intelligent Systems, EarlyView.
An instance‐level, model‐agnostic explanation of class differentiation is introduced through SHAP‐LCD, linking probability shifts to feature‐wise Shapley contributions. The method operates on tabular and image data and is released in a fully reproducible implementation, offering a transparent way to examine, at each instance, why predictive models ...
Roxana M. Romero Luna, Hugo A. Ordoñez Erazo, Carlos A. Cobos Lozada   +2 more
wiley   +1 more source

Time‐Delayed Spiking Reservoir Computing Enables Efficient Time Series Prediction

Advanced Intelligent Systems, EarlyView.
This study proposes time‐delayed spiking reservoir computing (TDSRC) for efficient time series prediction. By concatenating time‐lagged states, TDSRC constructs an expanded readout feature vector without altering internal reservoir dynamics. This approach enables highly accurate forecasting with significantly fewer neurons, providing a resource ...
Pin Jin, Ren Cao, Zhongxue Gan, Siyang Leng   +3 more
wiley   +1 more source

Design and Implementation of a Three-Layer Backpropagation Neural Network for Multi-Output Regression in Citizen-Science Impact Assessment

AI
Measuring the impact of citizen-science projects is hard because inputs are heterogeneous, mostly categorical, and sparse. We present Alquimics, a compact supervised neural network trained on one-hot project descriptors to predict impacts across five ...
Luigi Ceccaroni, Lyle Visa, Iain Visa
doaj   +1 more source

ParamNet: A Physics‐Guided Deep Learning Framework for Intelligent Self‐Inversion of Vacuum Optical Levitation Systems

Advanced Intelligent Systems, EarlyView.
A physics‐guided deep learning framework, ParamNet, is introduced for the intelligent self‐inversion of vacuum optical tweezers. By fuzing dual‐branch time–frequency features with physical dynamical constraints, it achieves high‐accuracy calibration of trap parameters from short‐window, low‐frequency trajectories, outperforming traditional methods ...
Qi Zheng, Wenfeng Fan, Feng Li, Jian Wu, Wei Quan   +4 more
wiley   +1 more source

Exploiting Ferroelectric and Spintronic Dynamics for Neural Network Computation

Advanced Intelligent Systems, EarlyView.
Ferroelectric and spintronic devices, relying on the control of polarization and magnetization, offer intrinsically fast, durable, energy‐efficient, and low‐latency building blocks for analog in‐memory computing. The hysteretic dynamics of an order parameter are leveraged to provide nonvolatile, multistate memory and nonlinear switching. Brain‐inspired
Dashiell Harrison, Kevin Nadaud, Josquin Harrison, Guillaume F. Nataf, Laura Bégon‐Lours   +4 more
wiley   +1 more source