High Variance Clue from Learning Curves
When training error is already near the best possible level, there is little opportunity to reduce bias. If dev error remains much worse than training error, the main room for improvement is reducing variance.
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
Reading a High-Bias Learning Curve When the Train-Dev Gap Is Small
Interpreting Training and Dev Error When Both Fit and Stability Are Poor
Why plot training error and development error on the same graph?
True or False: Looking only at the error from the largest training sample gives the full story of how performance changes as data increases.
Generalization Curves
Reading a Learning Curve
How to build and read a learning curve for a model
Why a full learning curve is more informative
Choosing a plot to judge whether more labeled data is worth the effort.
Why plot training and validation error together?
What does the far-right point on a learning curve indicate?
One data point is enough to understand how a model will change as more training examples are added.
Learn After
What learning-curve pattern usually indicates high variance?
When a model's training error is already near the best attainable error, there is usually still a large amount of bias reduction left.
Adding _____ to Improve Generalization
Match each learning-curve idea to its meaning.
Diagnosing a High-Variance Learning Curve
Reading Bias and Variance from Error Rates
Interpreting a Wide Train-Validation Gap
Why Overfitting Leaves More Room for Variance Reduction
What is the most likely effect of adding more training examples in a high-variance setting?
A Model Can Fit Training Data Well Yet Still Perform Poorly on New Data