The error is the object to be optimized, and the neural network is set to minimize the error. Take an example: for example, give a bunch of independent variables x1, X2, X3, x4, and a dependent variable Y1. In the training center. One set of data is [2,4,6,7] and 12. The output of neural network is 11. Then the error is 11-12 = - 1. Error is "actual output network output". The objective function should minimize the error to achieve the goal of accurate prediction.<br>
正在翻译中..