Most normalization methods map the range of a single attribute to another range, typically 0 to 1 or -1 to +1.
Normalization is very sensitive to outliers. Without outlier treatment, most values will be mapped to a tiny range, resulting in a significant loss of information. (See"Routines for Outlier Treatment".)
Table 4-5 Normalization Methods in DBMS_DATA_MINING_TRANSFORM
| Transformation | Description |
|---|---|
|
This technique computes the normalization of an attribute using the minimum and maximum values. The shift is the minimum value, and the scale is the difference between the maximum and minimum values. |
|
|
This normalization technique also uses the minimum and maximum values. For scale normalization, shift = 0, and scale = max{abs(max), abs(min)}. |
|
|
This technique computes the normalization of an attribute using the mean and the standard deviation. Shift is the mean, and scale is the standard deviation. |