August 2019
Intermediate to advanced
342 pages
9h 35m
English
One way to intuitively understand the different tasks associated with individual NNs involved in a GAN is to consider the scenario in which the discriminator network tries to correctly classify spam messages artificially generated by the generator network. To demonstrate the different objective functions that the individual NNs must optimize, we will resort to conditional probabilities (which are the basis of the Bayes' rule), which we have already encountered in Chapter 3, Ham or Spam? Detecting Email Cybersecurity Threats with AI, in the section Spam detection with Naive Bayes.
We define P(S|W) as the probability that a given email message represents spam (S), based on the presence within the text ...
Read now
Unlock full access