Prediction of Eye State Using KNN Algorithm

In this research paper, basic machine learning methodology for the classification of Eye State (i.e., Eyes Open or Closed) using Electroencephalography (EEG) Data is suggested. The idea is to compare and validate that basic Machine Learning (ML) approach (K-Nearest Neighbors KNN) can also provide better prediction accuracy in certain domains (in this case eye state prediction) than complex ML approaches (Support Vector Machine (SVM), Artificial Neural Network (ANN), or Deep Neural Network (DNN). The EEG data was collected using EMotiv EPOC headset and each record was labelled manually, containing 14 channels (columns of the record) using camera as open or closed eyes. The experimental results validate that stated methodology of using KNN provides better prediction accuracy in lesser time than other complex ML approaches.