At each training iteration, the descriminator is trained like a normal binary classifier, then the generator is trained to maximize the discriminator's error.TrueFalse