Using Public Web Images When No Better Future-Like Data Exists
If no data source closely resembles the inputs the system will face after deployment, a team may begin with public web images. That choice can be reasonable, but it carries the risk that the resulting model will not work well on real-world inputs that differ from the training set.
0
1
Tags
Machine Learning
Deep Learning
Machine Learning Strategy
Supervised Learning
Dive into Deep Learning @ D2L
Data Science
Machine Learning Yearning @ DeepLearning.AI
Related
Some Data Should Be Left Out of Training
When Development and Test Sets Reflect Different Populations
How Model Capacity Changes the Risk of Mixing Data Sources
One Predictor Can Work Across Multiple Data Sources
Choose evaluation data to match the real-world target
Mismatched Auxiliary Data Source
Building Dev and Test Sets Before Real Users Exist
Refreshing Evaluation Sets After a Product Launch
Using Public Web Images When No Better Future-Like Data Exists
Judging How Much to Invest in Dev and Test Sets
What should determine dev and test set selection?
True or False: You can assume the training set and test set always come from the same distribution.
Development and test sets should reflect the conditions you expect after deployment, not only the _____ available in your training pool.
Why can a simple random test split be a poor choice when the data you expect in the future is different from the data you have now?
You can usually assume the data used for training and the data used for testing come from the same distribution.
Design Dev and Test Sets for the Future
Match each concept about development and test sets to its description.
Order the steps for choosing development and test sets when future data differs from training data.
What should dev and test examples be designed to resemble?
A validation and test set must exactly match the training distribution in every project.
How should a test set be chosen when deployment data will differ?
Match each data scenario to the best dev/test set choice.
Order the reasoning steps for deciding whether a dev/test split is appropriate.
Why a Random 30 Percent Split Can Be Misleading When Future Data Will Differ
Dev and Test Splits for a Field-Photo Classifier
How to Choose Dev and Test Data When Future Data Will Differ
Learn After
When is it reasonable to use web-scraped images as a fallback training source?
True or False: Training a model using only images from a polished company website guarantees good performance on real-world images.
A model trained only on one narrow data source may not _____ when conditions change.
Match each data situation to its associated risk or outcome.
Order the decision steps for using website photos when future-like data is hard to obtain.
Weighing proxy photos when no future-like data exists
Explain a performance gap between catalog photos and live camera images.
What risk comes from training on web-scraped photos when the deployment data will differ?
What is the best interpretation of using website images as training data?
True or False: Website images should be your first choice for gathering training data.