#representation-learning

7 questions from Neural Network Basics, Computer Vision Basics, Autoencoders & VAEs

Timed test

7 questions · 7 min · score and explanations at the end

Q 1 / 7Score 0 / 0

Shuffle questions

Question 1: Single choice

easyNeural Network Basics

Why do deep networks need much less hand-engineered feature design for images?

Once a network has enough layers it no longer needs training data.1
Each layer applies features that experts fixed before training.2
Each image is reduced to its average pixel value.3
Their layers learn useful feature representations from the training data.4

Keys: 1–9 select · Enter check / next