Z-normalization means rescaling the feature $X$ by subtracting the average $\mu$ and dividing by its standard deviation $\sigma$, i.e., $(X-\mu)/\sigma$.
What is the usefulness of normalizing data for machine learning?
Asked
Active
Viewed 155 times
0
-
You may wanna look at this for implications of normalization in different ML tasks http://sebastianraschka.com/Articles/2014_about_feature_scaling.html – Zoran Jul 12 '14 at 11:16
-
One point that the link does not mention is the representation of continuous numbers using discrete ones. there is denser representation around 1. http://en.wikipedia.org/wiki/IEEE_floating_point – EngrStudent Jul 12 '14 at 14:49