May 2020
Intermediate to advanced
530 pages
17h 8m
English
In practice, mini-batches are often used as a compromise. The weights are adjusted after processing several training samples (mini-batches). This is done less often than with stochastic descent, but more often than with batch mode.
Now that we've looked at the main backpropagation training modes, let's discuss the problems of the backpropagation method.
Read now
Unlock full access