A NEW FRAMEWORK BASED ON EXTREME LEARNING MACHINE FOR EPILEPTIC SEIZURE DETECTION
Parthiban Kg, E. Saranya, S. Vijayachitra, N. Gomathi · 2014
A sharp cause for the seizure remains within the darker aspect of the detection. To develop correct realizable automatic spike detection improvement methodology has been projected. The reliable application of machine learning strategies becomes progressively vital in difficult engineering domains. especially, the applying of Extreme Learning Machines (ELM) looks promising attributable to their apparent simplicity and therefore the capability of terribly economical process of enormous and high)dimensional knowledge sets. However, the ELM paradigm is predicated on the conception of single hidden)layer neural networks with arbitrarily initialized and glued input weights and is therefore inherently unreliable. The goal is to produce the Extreme Learning Machine approach with the talents to perform dependably in numerous, difficult engineering tasks by exploiting the simplicity, catholicity and procedure potency of the model.