4. Errors

How are errors in artificial learning minimised?

In the previous units, we have already come across the term learning or training algorithm. This chapter focusses on how this algorithm manages to minimise the error of a neural network. To do this, we can play the trainer ourselves by taking on the task to minimise the error of the network. We have some example data sets at our disposal, which can be worked through one by one with increasing level of difficulty: simple, simple with help and difficult.