CODEWITHSUNDEEP
pythonmachine-learningscikit-learn
Richardsop
Richardsop

Reputation: 53

Proper way of creating feature space for Machine learning algorithms

This question is maybe general. Can a feature space consisting of negative feature values be used train a machine learning algorithm i.e., Logistic regression?

feature,feature,feature,class-label  
-0.5,0.2,1,yes         
0.9,-0.7,2,no   
 ..........     
...........

Upvotes: 0

Views: 215

Answers (1)

lejlot
lejlot

Reputation: 66835

Feature space has to be consistent with given model. Most of the machine learning classifiers (including Logistic regression) require features to be:

  • real numbers (sign does not matter)
  • constant length

and that's all.

Upvotes: 2

Related Questions