Results 181 to 190 of about 2,004,297 (387)
Flowsheet generation through hierarchical reinforcement learning and graph neural networks
AIChE Journal, Volume 69, Issue 1, January 2023., 2023 Abstract Process synthesis experiences a disruptive transformation accelerated by artificial intelligence. We propose a reinforcement learning algorithm for chemical process design based on a state‐of‐the‐art actor‐critic logic. Our proposed algorithm represents chemical processes as graphs and uses graph convolutional neural networks to learn from ...
Laura Stops +3 more
wiley +1 more source
Cascaded ensemble of convolutional neural networks and handcrafted features for mitosis detection [PDF]
, 2014 Haibo Wang +8 more
openalex +1 more source
Progress in Surface Plasmon and Other Resonance Biosensors for Biomedical Applications
Advanced Materials Technologies, EarlyView.This is the shortened version: Recent advancements in surface plasmon resonance and other optical resonance biosensors for biomedical applications are presented. Advanced sensing strategies are examined for the detection of diverse analytes, integration of nanomaterials and machine learning, and emerging nonplasmonic modes like guided mode resonance ...
Faten Bashar Kamal Eddin +8 more
wiley +1 more source
Head-pose invariant facial expression recognition using convolutional neural networks
, 2003 Beat Fasel
openalex +2 more sources
Advanced Optical Materials, EarlyView.Photochromic compounds are versatile ingredients for the development of Chemical AI. When they are embedded in a tight microenvironment, they become Markov blankets. They are also valuable for processing Boolean and Fuzzy logic. They contribute to neuromorphic engineering in wetware based on opto‐chemical signals exchanged with oscillatory chemical ...
Pier Luigi Gentili
wiley +1 more source
Twin neural network regression
Applied AI Letters, Volume 3, Issue 4, December 2022., 2022 We propose to reformulate a regression problem into predicting differences between target values. This allows for leveraging consistency conditions which can be used as uncertainty estimates and enable the production of an ensemble of predictions while training only a single neural network.
Sebastian Johann Wetzel +3 more
wiley +1 more source
Efficient Training Algorithms for a Class of Shunting Inhibitory Convolutional Neural Networks
, 2005 Fok Hing Chi Tivive +1 more
openalex +2 more sources
Advanced Photonics Research, EarlyView.Ultrafast pulsed laser technology enables precise material processing. This study examines molybdenum oxide formation under varying laser parameters. A dataset of 187 samples is generated, and eight deep neural networks (DNNs) with different architectures are trained. Models are validated using three learning rates and evaluated with mean squared error,
Jose R. Paredes‐Miguel +6 more
wiley +1 more source
Analysis of the rate of convergence of an over-parametrized convolutional neural network image classifier learned by gradient descent [PDF]
arXivImage classification based on over-parametrized convolutional neural networks with a global average-pooling layer is considered. The weights of the network are learned by gradient descent. A bound on the rate of convergence of the difference between the misclassification risk of the newly introduced convolutional neural network estimate and the minimal
arxiv
Advanced Photonics Research, EarlyView.A novel free‐space optical communication protocol that combines the benefits of partially coherent light sources with the decoding capabilities of convolutional neural networks is proposed. This approach establishes a robust communication link that can withstand atmospheric turbulence.
Linxuan Yao +6 more
wiley +1 more source