Detection of Ocular Artifacts Using Bagged Tree Ensemble Model

Removal of artifacts from the biomedical signal is a cumbersome task. As the desire of physicians is a clean signal for diagnosis authors have tried to detect the artifacts within the signal. Detection of artifacts is a primary job to remove them with further techniques. In this paper we have collected data from Mendeley database and focused on the ocular artifacts. The ensemble method is chosen by the method of bagging and boosting to enhance the detection accuracy. As the technique is one of the statistical techniques, it is found better accuracy in this work. From the dataset 19 channel Ocular artifactual signals are considered along with the healthy signal. The features have been extracted using time domain and frequency domain techniques. Finally, the combination with ensemble classifier shows better accuracy as explained in the result section.