An attribute has a missing value if the value is NULL
.
Some data mining algorithms, such as Support Vector Machine, are sensitive to missing values in data; for more information, see Missing Values in Oracle Data Mining.
Note: Do not use the missing values wizard to treat sparse data (data for which only a small fraction of the attributes are non-zero or non-NULL
for any given case). Sparse data is typically found in market basket applications. For more information, see Sparsity.
The Missing Values Transformation wizard identifiers attributes with missing values and lets you specify how to treat the missing values. The wizard provides several ways to treat missing values: replacing the missing value with some other value (minimum, maximum, mean, mode, or a custom value), dropping the attribute from the view, and dropping those cases where the attribute is NULL
.
Alternatively, you can generate default missing values treatments by specifying the algorithm to be used for model build.
A default strategy for missing values is provided if the wizard is launched from a Mining Activity.
This transformation processes single-record case and multi-record case data, both categorical and numerical.
If you do not want to see this page the next time that you launch the wizard, check the Skip this Page Next Time box.
Click Next to continue.
Copyright © 2006, 2008, Oracle. All rights reserved.