Neural networks are computational models composed of interconnected processing units, or artificial neurons, organized in layers that transform input data through weighted connections and nonlinear activation functions. Inspired by biological nervous systems, they learn by adjusting connection weights to minimize er…
Neural networks are computational models composed of interconnected processing units, or artificial neurons, organized in layers that transform input data through weighted connections and nonlinear activation functions. Inspired by biological nervous systems, they learn by adjusting connection weights to minimize error, enabling pattern recognition, classification, regression, and prediction across complex, high-dimensional data. Architectures range from shallow feedforward networks to deep and convolutional models, and training methods such as backpropagation and transfer learning underpin their performance. Research in this area applies neural networks and related machine-learning methods to diverse scientific and engineering problems, including genetic-algorithm-coupled networks for estimating subsurface features of the Earth, artificial neural network models for analysing long-term rainfall data and climate-change patterns, and nature-inspired optimization for interpreting geoelectrical data. Applications in agriculture include automated grassweed detection in wheat systems and deep-learning and transfer-learning approaches for detecting crop-leaf diseases, while biomedical applications include dynamic network analysis of functional connectivity in dementia and non-invasive measurement techniques. Time-series modelling, such as seasonal autoregressive models for pandemic prediction, illustrates the breadth of predictive use. By learning representations directly from data, neural networks complement and extend traditional statistical and physical models. The journal publishes peer-reviewed research on neural-network methods and their application within applied artificial intelligence and across scientific domains.
12 peer-reviewed articles, ranked by relevance. Each links to its DOI.
The 12 articles above have been cited 141 times in the scholarly literature. Citation data via OpenAlex and Crossref, updated Jun 2026.
A sample of recent works citing this journal's research on Neural Networks, linking to each citing work.