#data-augmentation

14 questions from Training Deep Networks, Convolutional Neural Networks, Transfer Learning, Computer Vision Basics, Object Detection, Segmentation

Timed test

14 questions · 14 min · score and explanations at the end

Q 1 / 14Score 0 / 0

Shuffle questions

Question 1: Select all that apply

easyAI-checkedTraining Deep Networks

Besides weight decay, more training data and a model of suitable size, which of these help against overfitting? Select all that apply.

Early stopping based on the validation error1
Data augmentation, such as random crops and flips of the training images2
Training for many more epochs, until the training loss reaches zero3
Adding higher-degree polynomial features of the inputs4
Dropout on the hidden layers5

Keys: 1–9 select · Enter check / next