Twenty Eyeball Dev Set Errors Give Rough Error-Source Sense
If a classifier makes about 20 mistakes on Eyeball dev examples, manual review starts to give a rough sense of the major error sources.
0
1
Tags
Machine Learning
Deep Learning
Machine Learning Strategy
Supervised Learning
Dive into Deep Learning @ D2L
Data Science
Related
Ten Eyeball Dev Set Errors Are Very Small but Useful
Twenty Eyeball Dev Set Errors Give Rough Error-Source Sense
Fifty Eyeball Dev Set Errors Give Good Error-Source Sense
One Hundred Eyeball Dev Set Errors Give Very Good Error-Source Sense
Eyeball Dev Set Size Depends on Classifier Error Rate
What is the primary goal of ensuring the Eyeball dev set is large enough?
The rough size guidelines for the Eyeball dev set in ML Yearning apply specifically to tasks that humans do well.
The Eyeball dev set should be large enough to give a sense of the algorithm's major _____ categories.
Match each Eyeball dev set misclassified-example count to its described level of insight.
Order the steps for using an Eyeball dev set to surface major error categories.
Which example task does Andrew Ng cite when introducing rough Eyeball dev set size guidelines?
A very small Eyeball dev set can reliably reveal all of an algorithm's major error categories.
Andrew Ng's rough Eyeball dev set size guidelines apply when _____ perform the task well.
Match each Eyeball dev set concept to its correct definition in ML Yearning.
Order the reasoning steps for deciding whether an Eyeball dev set is large enough.