Learn Before
When Abundant Labels for Substeps Support a Staged Design
If you have a large labeled dataset for the intermediate parts of a system, it can make sense to build the solution in stages. The extra data can be used to train each intermediate component well, which may make a multi-stage pipeline more practical and effective than trying to learn everything in one model.
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
When Abundant Labels for Substeps Support a Staged Design
What factor should guide how you divide a modular prediction system into components?
Labeled data access can shape component choices in a modular system.
Choose components where training data can be collected _____.
Match each modular-system term with its meaning in pipeline planning.
Order the reasoning used to decide which pipeline modules should be trained from data.
Why training-data availability should affect pipeline component choice
Select a pipeline component using training-data availability.
Checking Data Readiness for a Pipeline Step
Which option best fits a component that needs easy-to-gather training examples?
Training Data Availability Should Be Checked When Choosing a Component
Learn After
Why Labeled Perception Data Can Make a Staged Driving System Easier
When is a staged model design especially attractive?
A staged design can be better when each stage has usable training data.
Available data for _____ in a staged workflow
Match staged system concepts
Choosing a Multi-Stage Pipeline When Subtask Labels Are Abundant
When a Modular Pipeline Can Be Competitive
Choose a pipeline structure when intermediate labels are abundant
When can a staged pipeline outperform a single direct model?
Which example is an intermediate module in a larger machine learning pipeline?
Limited data makes a multi-stage pipeline the best choice.