Saved Bookmarks
| 1. |
What Is Data Normalization And Why Do We Need It? |
|
Answer» Data normalization is very important preprocessing step, USED to rescale values to fit in a specific range to assure better CONVERGENCE during backpropagation. In GENERAL, it boils down to subtracting the mean of each data point and DIVIDING by its standard deviation. Data normalization is very important preprocessing step, used to rescale values to fit in a specific range to assure better convergence during backpropagation. In general, it boils down to subtracting the mean of each data point and dividing by its standard deviation. |
|