An Approach towards Evaluation of EEG Dataset for Sleep Data Analysis

Bio-informatics becomes more popular now-a-days as to combine medical problems with the information technology technical field to identify or to detect the medical disorders in the human body. This field is valuable for both the medical practitioners as well as software developers. To identify the brain disorders like epilepsy, parasomnias in the patients, EEG data is fetched from physionet data source where numbers of datasets are available for researchers. The analysis of this dataset is done by polyman which is used as plotting libraries and transformation tool for transforming the EDF data format into ASCII data format. The tool also helps in identifying the different parameters which are useful for predictions related to brain disorders. In this work, improved algorithm Support Vector Regression (SVR) has been developed and implemented using advanced python programming language and compared with classical Support vector Machine (SVM) algorithm. Results show better performance of SVR than SVM in terms of both complexity and execution time.