Icon

01_​Combining_​Classifiers_​using_​Prediction_​Fusion

How to use the Prediction Fusion node

This workflow shows how the prediction fusion node can be used to combine the predictions of a naive bayes and a svm classifier.

Use Prediction Fusion to combine predictions of different classifiers The Prediction Fusion node allows to fuse the predictions of two or more classifiers intoone prediction. This workflow shows how this is done with a Naive Bayes and a SVMclassifier. read workflow local datapartition data in 80% training20% testingusing stratified samplingtrain svm classifier usingrbf kernel with sigma 0.1predict class withsvmcombine predictionsusing the mean of the predictionsjoin the prediction tablesonly keep the class columnand prediction columnsscore prediction ofnaive bayesscore fusedpredictionscore predictionof svmtrain naive bayes classifierpredict class withnaive bayes File Reader Partitioning SVM Learner SVM Predictor Prediction Fusion Joiner Scorer Scorer Scorer Naive Bayes Learner Naive BayesPredictor Use Prediction Fusion to combine predictions of different classifiers The Prediction Fusion node allows to fuse the predictions of two or more classifiers intoone prediction. This workflow shows how this is done with a Naive Bayes and a SVMclassifier. read workflow local datapartition data in 80% training20% testingusing stratified samplingtrain svm classifier usingrbf kernel with sigma 0.1predict class withsvmcombine predictionsusing the mean of the predictionsjoin the prediction tablesonly keep the class columnand prediction columnsscore prediction ofnaive bayesscore fusedpredictionscore predictionof svmtrain naive bayes classifierpredict class withnaive bayesFile Reader Partitioning SVM Learner SVM Predictor Prediction Fusion Joiner Scorer Scorer Scorer Naive Bayes Learner Naive BayesPredictor

Nodes

Extensions

Links