Error Factor

Top 

Error Factor

When working with supervised networks, each training pattern has a known output that you want the neural network to be able to reproduce.  When each training pattern is presented to the network, NeuroShell 2 computes the error between the actual outputs in the training pattern and the network's predictions for each of the network's outputs.  The total error for each pattern is the sum of the squares of the differences.  The measurements are made in NeuroShell 2's internal interval [0, 1] for speed.