Icon

ensemble models demo

Ensemble models demoData: Customer data (18000+ rows) with various numerical and categorical featuresGoal: Build a classifier for Target (0/1)1. Data is read from the table file2. Partitioning to the training data (70%) and testing data (30%)3. Decision tree model4. Random forest model5. Gradient boosted trees model Decision tree modelDecision Tree Learner & PredictorMinimum node size = 5Otherwise default setting Random forest modelRandom Forest Learner & PredictorMinimum node size = 5Number of models = 100Otherwise default setting Gradient boosted trees modelGradient Boosted Trees Learner & PredictorMaximum depth = 10Number of models = 100Otherwise default setting reading thedata tableTraining 70%Testing 30% Table Reader Partitioning DecisionTree Learner Decision TreePredictor Scorer Random ForestLearner Random ForestPredictor Scorer Gradient BoostedTrees Learner Gradient BoostedTrees Predictor Scorer Ensemble models demoData: Customer data (18000+ rows) with various numerical and categorical featuresGoal: Build a classifier for Target (0/1)1. Data is read from the table file2. Partitioning to the training data (70%) and testing data (30%)3. Decision tree model4. Random forest model5. Gradient boosted trees model Decision tree modelDecision Tree Learner & PredictorMinimum node size = 5Otherwise default setting Random forest modelRandom Forest Learner & PredictorMinimum node size = 5Number of models = 100Otherwise default setting Gradient boosted trees modelGradient Boosted Trees Learner & PredictorMaximum depth = 10Number of models = 100Otherwise default setting reading thedata tableTraining 70%Testing 30%Table Reader Partitioning DecisionTree Learner Decision TreePredictor Scorer Random ForestLearner Random ForestPredictor Scorer Gradient BoostedTrees Learner Gradient BoostedTrees Predictor Scorer

Nodes

Extensions

Links