A schematic diagram of a neural network with a double hidden layers of 3-3 neurons (H1 =3, H2 =3) and two input variables (I= 2). (IMAGE)

Nuclear Science and Techniques

A schematic diagram of a neural network with a double hidden layers of 3-3 neurons (H1 =3, H2 =3) and two input variables (I= 2).

Caption

This structure represents the typical multilayer perceptron network, also known as a "backpropagation" or "feedforward" network, which serves as the foundation for the Bayesian Neural Network (BNN) models employed to predict fragment yields in photon-induced fission reactions of thorium isotopes. The input variables are fed into the network, and after processing through the hidden layers using a nonlinear activation function, the output is produced. This diagram visually explains how the model parameters (biases and weights) are organized to map input data to output predictions.

Credit

machunwang@126.com

Usage Restrictions

Credit must be given to the creator. Only noncommercial uses of the work are permitted.

License

CC BY-NC

Disclaimer: AAAS and EurekAlert! are not responsible for the accuracy of news releases posted to EurekAlert! by contributing institutions or for the use of any information through the EurekAlert system.