Skip to content
ml-recall

#memorization

2 questions from Training Deep Networks, Generative Adversarial Networks

Timed test

2 questions · 2 min · score and explanations at the end

Take a test
Q 1 / 2Score 0 / 0

Question 1: Short answer

easyAI-checkedTraining Deep Networks

A large network is trained on an image dataset with 3 balanced classes (cat, dog, wild animal) after the labels were randomly shuffled, so labels are unrelated to images. It reaches 100% training accuracy. About what validation accuracy, in percent, should you expect on correctly labelled validation images?

a percentage, e.g. 50