How Model Capacity Changes the Risk of Mixing Data Sources
Combining photos from two different sources, such as employee-submitted pictures and web-sourced images, could hurt earlier vision systems built from hand-crafted features and a simple linear classifier. With modern, highly flexible neural networks, that kind of harm is much less likely.
0
1
Tags
Machine Learning
Deep Learning
Supervised Learning
Dive into Deep Learning @ D2L
Data Science
Machine Learning Strategy
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
Combining Two Image Sources with a High-Capacity Model
Why is it less risky today to combine customer-supplied photos with web-scraped photos in one training set?
True or False: Combining data from different sources could sometimes make older machine-learning systems perform worse.
A classic early vision pipeline used hand-crafted image features followed by a simple _____ classifier.
Match each model type or factor to its role in the risk of combining training data sources.
Order the reasoning steps for deciding whether to combine training datasets.
Why model flexibility changes the risk of combining training data sources
Should this team combine two review datasets with a simple model?
What makes combining training data from multiple sources risky?
Which model family is most exposed to trouble when two data sources disagree?
True or False: Using a very large neural network completely removes the risk of problems when combining multiple training data sources.