#non-stationary

Timed test

3 questions · 3 min · score and explanations at the end

Q 1 / 3Score 0 / 0

Shuffle questions

In which training situation is the objective non-stationary?

A fraud detector is retrained continually while fraudsters keep changing tactics1
An image classifier is trained for many epochs on a fixed dataset, shuffled anew every epoch2
A network is trained on a fixed dataset whose loss surface has many local minima3
The learning rate of a model trained on a fixed dataset is halved every ten epochs4

Keys: 1–9 select · Enter check / next