Glossary term
What is Data leakage?
In Apex Flow Academy, "Data leakage" means: Information from the test rows sneaking into training or preparation, which makes the score look better than it really is. It is taught in Machine Learning With Python: From Zero to a Working Model. Related terms include Validation set, Impute, Test set and Median.
- 1 course teaches it
- plain-word definition
The facts
What is Data leakage?
Information from the test rows sneaking into training or preparation, which makes the score look better than it really is.
Where is Data leakage taught?
Data leakage is defined in Machine Learning With Python: From Zero to a Working Model.
Which terms are related to Data leakage?
Next to Data leakage in the course glossaries: Validation set, Impute, Test set, Median, Training set, One-hot encoding, Accuracy and Standardising (scaling).
Validation setImputeTest setMedianTraining setOne-hot encodingAccuracyStandardising (scaling)
The hard questions, answered straight
Bold questions people really ask. Each answer comes from the course data or a settled Academy fact.
Do I need experience before I learn Data leakage?
It depends on the course. Each course that teaches it lists its own prerequisites, and all are listed at beginner level:
- Machine Learning With Python: From Zero to a Working Model: Basic Python: variables, lists, dictionaries, functions, and running a .py file from a terminal.