Brain Tumor Classification using Ensemble Classifiers

In Medical analysis and treatment the primary step is to identify the presence of tumor in brain and the effected area. In this regard, Brain Tumor (BT) classification helps the doctor to know the stage of tumor. This paper presents BT classification using a set of ensemble classifiers (EC). To train the classifier 17 redundant noise features are extracted from the MRI images. In training phase, these features are fed to variety of EC classifiers for learning. In testing phase, the trained model is used to identify the class of MRI image. Performance of proposed EC classifiers are analyzed under different training rates. Performance is analyzed for different MR image data sets under different training rates. Experimental results depicted that EC classifiers performance is superior than most of the traditional machine learning (ML) classifiers such as random forecasting, logistic regression, and naive bayes classifiers.