Decision Tree Builder
Decision trees are a predictive analytics visualization used to evaluate visitor characteristics and relationships. The Decision Tree Builder generates a decision tree visualization based on a specified positive case and a set of inputs.
A Decision Tree is a binary classifier with a set of rules (or filters) identifying visitors who satisfy specific rules based on a positive case. A decision tree sets rules to classify visitors who satisfy (or do not satisfy) this positive case. These rules generate a tree map to provide a level of confidence to meet these positive case results.
A Decision Tree is built by examining inputs at each level and choosing the one that provides a maximum gain of information at a specified split point. Split points for each variable-level generates two sets:
- Values less than or equal to the split point, and
- Values greater than the split point.
Use decision trees to
- Perform meaningful analysis and interpretation in less time.
- Employ automated segment generation.
- Quickly make inferences from a model based on a large amount of data.