Artificial neural networks

Machine learning

Advances in Nanotechnology: Predicting Thermophysical Properties of Nanoparticles

Researchers at Shahid Bahonar University of Kerman in Iran have made significant progress in understanding the thermophysical properties of alumina nanoparticles in binary mixtures of water and ionic liquids. Using advanced machine learning algorithms, specifically Cascaded Forward Neural Networks (CFNN) and Generalized Regression Neural Networks (GRNN), the team has successfully

Machine learning

Breakthrough in Environmental Monitoring: Novel Machine Learning Approach for Anomaly Detection

Researchers from the National Center for Radiation Research and Technology have made a significant contribution to environmental monitoring with their novel machine learning approach for anomaly detection in gamma-ray spectra. This innovative technique combines neural network modeling with bio-inspired optimization, enabling the identification of anomalies even at low source to

Machine learning

Revolutionary Integration of Machine Learning and Vibrational Spectroscopy for Food Safety

The integration of machine learning with vibrational spectroscopy has significantly advanced the analysis of food quality, authenticity, and safety. Ohio State University researchers reported that the synergy between machine learning and vibrational spectroscopy methods, including near-infrared, mid-infrared, and Raman spectroscopy, has enhanced capabilities for identifying adulterants, quantifying quality indicators, and

Artificial intelligence

AI-Powered Diagnostic System for High-Grade Prostate Cancer Shows Promising Results

Researchers from Chiba University, Japan, have developed an artificial intelligence (AI) powered diagnostic system for high-grade prostate cancer using innovative 3D magnetic resonance imaging (MRI) sequences and a 3-dimensional deep convolutional neural network (3DCNN). The study, published in the journal Current Urology, aimed to develop a diagnostic support system for

Artificial intelligence

Hybrid Recurrent Neural Network Architecture for Subscriber Traffic Prediction

Research in the field of applied computational intelligence and soft computing has led to the development of a new hybrid recurrent neural network architecture for predicting subscriber traffic in mobile telecommunication networks. The proposed architecture combines features of Jordan and Frasconi networks and leverages incremental learning through restricted Coulomb energy

Machine learning

Artificial Neural Networks Improve Railway Shunting Route Search Accuracy and Reduce Memory Consumption

Researchers from Shandong Jiaotong University have proposed an automatic railway shunting route search model based on an improved artificial neural network algorithm to enhance route search accuracy while reducing memory consumption. The method constructs a time-varying railway network topology through dynamic topology modeling and conflict avoidance mechanisms, abstracting signals and

Machine learning

Efficient Real-Time Nodal Demand Forecasting in Water Distribution Systems

Researchers from Zhejiang University have proposed a novel attention-augmented gated graph neural network (AGN) for real-time demand forecasting in water distribution systems. The AGN model overcomes limitations of convolution-based graph neural networks by capturing long-range dependencies and dynamic node interactions, leading to enhanced performance on real-world modeling problems. The study

Artificial intelligence

Voronoi Diagram-Based Sampling Methods Improve Prediction Accuracy in Physics-Informed Neural Networks

Researchers have developed novel sampling methods based on Voronoi diagrams to improve the performance of Physics-Informed Neural Networks (PINNs) for solving partial differential equations (PDEs). These methods, which consider the locations of generated points and the domain each point covers, have been tested in six different PDE simulation experiments and

Artificial intelligence

Breakthrough in Optical Computing for Artificial Intelligence

Research at Koc University has made significant strides in developing an innovative method for optical computing in artificial intelligence. The study, titled "Genetically programmable optical random neural networks," presents a novel approach to achieving high performance in optical neural networks with a simple and scalable design. The method

Machine learning

Machine Learning Breakthrough in Molecular Dynamics Simulations

Researchers from the Ames Laboratory, led by Weiyi Xia, have made a significant breakthrough in machine learning, developing an accurate and transferable artificial neural network machine learning (ANN-ML) interatomic potential for molecular dynamics simulations of the La-Si-P system. This study demonstrates the crucial role of training data in ML potential

Artificial intelligence

Researchers Develop AI Framework to Quantify Traditional Chinese Medicine Mechanisms

Researchers from China Pharmaceutical University have developed a novel interpretable graph artificial intelligence (GraphAI) framework to study the complex compatibility mechanisms of traditional Chinese medicine (TCM). The study, published in the Journal of Pharmaceutical Analysis, features a multidimensional TCM knowledge graph that integrates various standardized modules, including TCM terminology, Chinese

Artificial intelligence

Breakthrough in Artificial Intelligence: Nonreciprocal Neural Networks for Decoupled Bidirectional Analog Computing

Researchers from Zhejiang University have made a significant discovery in the field of artificial intelligence, introducing a nonreciprocal neural network that leverages enhanced magneto-optical effects to decouple forward and backward paths in computing. This innovation enables the creation of integrated perception-response systems, which are essential for various applications, including image