It is important to know the weight and threshold for what we want to allow computers to recognize the value of uncertainty. You can set your computer to adjust those values by teaching it to recognize patterns (Pattern) of what we want to give it recognition, called "Back propagation", which is the reverse process of recognition for training feed-forward. neural networks are used by the algorithm. Back-propagation In order to improve the capacity of the network (network weight) after the data model for the training of the network at a time, then the value received (output) of the network is compared to the expected results. Then calculate the error. The margin of error is sent back into the network to modify the weights the next.
การแปล กรุณารอสักครู่..
