The Predictive Scoring setting view provides a Feature Guess function. When you click Guess Columns, Treasure Data suggests highly relevant features (predictors) derived from attributes:
The Feature Guess function:
- Sends a request to API server
- Samples 100,000 values from every customer attributes in an audience
- Applies pre-defined set of rules for each attribute
Attributes detected as meaningless are automatically dropped, and pairs of a type (quantitative, categorical, or categorical array) and preprocessing rules (such as, how to fill in missing values, extract day of week from the timestamp) for potentially informative attributes are suggested in the input boxes.
The pre-defined set of guessing rules are based on:
- attribute type (number, string, array)
- attribute name
- regular expression
- standard deviation
of the sampled values.