1
vote
Is it possible for a feature not correlated with a dependent variable to become important in a machine learning model?
In the example above, the x1 and x2 features have extremely low correlation with the color variable, both being less than ...
- 4,044
1
vote
Is it possible for a feature not correlated with a dependent variable to become important in a machine learning model?
Yes. Perhaps the simplest examples have no linear relationship between the feature and target and hence zero correlation in the usual statistical sense, but have a clear nonlinear relationship (e.g. $...
- 12k
Only top scored, non community-wiki answers of a minimum length are eligible