Skip to content
free
K
1
2
practice
Home
Math
English
Science
Social Studies
Contact
Language
English
Español
Sign In
/
High School Electives
/
Computer Science
/
Overfitting and Generalization
Grade Level
Computer Science
HSCS.227 Overfitting and Generalization
0%
Skill Level
🔴 Hard
00:00
Progress to Next Level
25% to Medium
Easy
Medium
Hard
Mastery
100%
A model still gets 93% validation accuracy after the training labels are randomly shuffled. What should the team investigate first?
A
A breakthrough model that can learn without labels.
B
A need to train for more epochs on the shuffled labels
C
Underfitting caused by too much regularization.
D
Data leakage, duplicate records, or an evaluation bug that reveals the label indirectly.
Submit Answer