Figure 2
The diagram shows a neural network framework for forward and backward propagation. Five inputs labeled as Input 1 to Input 5 connect to a hidden layer through various weights denoted as W1 to W18. The forward propagation section includes functions for summation and activation, alongside a bias component. The output is linked to a prediction value and the true value, culminating in the calculation of a loss score. The backward propagation section indicates the process of updating weights and repeating the actions. Visual cues demonstrate the workflow from inputs to output and loss calculation. The layout organizes components logically, facilitating understanding of the network's operation.Schematic illustration of feed forward neural network with backpropagation algorithm
Source: Author’s own work