Icon

HR attrition prediction (baseline model)

There has been no title set for this workflow's metadata.

There has been no description set for this workflow's metadata.

HR ATTRITION PREDICTION Scoring new data! Develop here your main analysis (wrap the nodes into metanodes to safe space and be tidy!) Set the cut-off probabilitythat you want to use and overwrite thedefault prediction class column Model development dataset*hr_churn_dev_data"Scoring dataset"new_data_scoring" In case you performed data prep steps on traning data, do the same steps inside thismetanodeKeep only 4 columns:- ID- pred_prob-pred_class- group_numbername the output file"scoring_group_XX", changing XX with your group numberSubstitute this nodewith the predictor of yourselected model(to be connected with the correspondinglearner)Reconfigure this nodechanging the value setting:put your group number in place of 0075% trainingBase modelSave prob columnSelect custom threshold to optimize costsCalculatecostsAnalyze resultsRename the predictedattrition probability column "pred_prob"and the predicted class column"pred_class"Rule Engine Excel Reader Excel Reader Data prep Column Filter Excel Writer Logistic RegressionPredictor ConstantValue Column Partitioning LogisticRegression Learner Logistic RegressionPredictor Binary ClassificationInspector Math Formula Table View(JavaScript) Column Renamer HR ATTRITION PREDICTION Scoring new data! Develop here your main analysis (wrap the nodes into metanodes to safe space and be tidy!) Set the cut-off probabilitythat you want to use and overwrite thedefault prediction class column Model development dataset*hr_churn_dev_data"Scoring dataset"new_data_scoring" In case you performed data prep steps on traning data, do the same steps inside thismetanodeKeep only 4 columns:- ID- pred_prob-pred_class- group_numbername the output file"scoring_group_XX", changing XX with your group numberSubstitute this nodewith the predictor of yourselected model(to be connected with the correspondinglearner)Reconfigure this nodechanging the value setting:put your group number in place of 0075% trainingBase modelSave prob columnSelect custom threshold to optimize costsCalculatecostsAnalyze resultsRename the predictedattrition probability column "pred_prob"and the predicted class column"pred_class"Rule Engine Excel Reader Excel Reader Data prep Column Filter Excel Writer Logistic RegressionPredictor ConstantValue Column Partitioning LogisticRegression Learner Logistic RegressionPredictor Binary ClassificationInspector Math Formula Table View(JavaScript) Column Renamer

Nodes

Extensions

Links