Architecture different classifiers Classifier 1 level-2'features Classifier 2 F2 Combining Classifier Classifier 3 Classifier 4 Single feature set.different classifiers 15/55
Architecture - different classifiers 15 / 55
Classifiers Fusion o Fusion is useful only if the combined classifiers are mutually complementary Majority vote fuser:the majority should be always correct 16/55
Classifiers Fusion Fusion is useful only if the combined classifiers are mutually complementary Majority vote fuser: the majority should be always correct 16 / 55
Complementary classifiers o Several approaches have been proposed to construct ensembles made up of complementary classifiers.Some of them are: 1 Injecting randomness to the sample set 2 Varying the classifier type,architecture,or parameters 3 Manipulating training data 4 Manipulating features 17/55
Complementary classifiers Several approaches have been proposed to construct ensembles made up of complementary classifiers. Some of them are: 1 Injecting randomness to the sample set 2 Varying the classifier type, architecture, or parameters 3 Manipulating training data 4 Manipulating features 17 / 55
Outline (Level 1) Basic principle Multiple classifier combination Bagging Boosting 18/55
Outline (Level 1) 1 Basic principle 2 Multiple classifier combination 3 Bagging 4 Boosting 18 / 55