Icon

81606 - Applied Rule Identification

Decision Tree

Using the adult dataset, this workflow performs binary classification (income > or < 50K) using a Decision Tree. The target is the income column, either <=50K or <50K, predicted using the other demographic attributes. After partitioning the original dataset into training set and test, the decision tree is builkt on the training set and the final performance is evaluated on the test set using the Scorer node.

Extract individual rulesfrom rule set
Decision Tree to Ruleset
Read the adult dataset
CSV Reader
80% training 20% test stratified sampling
Partitioning
Train Decision Tree
Decision Tree Learner
Apply rules but instead ofassigning the Outcome weassign a Rule ID to each row
Rule Engine (Dictionary)
Keep RowID as aseparate column
RowID
Join Rule IDs with therespective rule to get themodel prediction*Outcome* is the prediction
Joiner

Nodes

Extensions

Links