A Survey: Evaluation of Ensemble Classifiers and pdf

A Survey: Evaluation of Ensemble Classifiers and

More Book Details

Description of the Book:

Over the past few decades, protein interactions have gainedimportance in many applications of prediction and datamining. They aid in cancer prediction and various otherdisease diagnosis. Imbalanced data problem in proteininteractions can be resolved both at data as well asalgorithmic levels. This paper evaluates and surveys variousmethods applicable at data level as well as ensemblemethods at algorithmic level. Cluster based under sampling,over sampling along with data based methods wereevaluated under Data level. Ensemble classifiers wereevaluated at the algorithmic level. Unstable base classifierssuch as SVM and ANN can be employed for ensembleclassifiers such as Bagging, Adaboost, Decorate, Ensemblenon‐negative matrix factorization and so on. Random forestcan improve the ensemble classification in dealing withimbalanced data problem over Bagging as well as Adaboostmethod for high dimensional data

• Book Topics/Themes: Bagging, Adaboost, Decorate, Oversampling, Under sampling

An excerpt captured from the PDF book