Approach for Solving Multiclass Problem of EEG Data using Multiobjective Simultaneous Learning Framework (AMEMS)

The Recent modern techniques, communication between humans and computers is proven a tremendous achievement in the field medical science. Computer hardware and signal processing have made possible the use of EEG signals or “brain waves” for Human-computer communication. Electroencephalography (EEG) is the electrical activity recording along the scalp. EEG refers to the recording of the spontaneous electrical activity in the brain over a short period of time, which is recorded from number of electrodes placed on the scalp. An Approach for solving multiclass problem of eeg data using multiobjective simultaneous learning framework (AMEMS), based on the concept of Multiobjective Particle Swarm Optimization (MOPSO) and Fuzzy Cmeans Clustering algorithm (FCM) is being analyzed and implemented. On the basis of implementation it is observed that, the training time and number of iterations achieved by MSCC classifier is comparatively higher. In this, the learning parameter is initially chosen at random to obtain the optimized clustering and classification performance. A new Approach for solving multiclass problem using multiobjective simultaneous learning framework is proposed which randomly initialized cluster center C to calculate learning parameter . With this approach, training time and number of iterations are reduced, thus significantly achieve improvement in performance of Clustering and Classification.