Comparative Analysis of KNN, SVM, DT for EOG based Human Computer Interface
- 1 September 2017
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 1023-1028
Abstract
In this study, EOG signal based human computer interface (HCI) has been implemented. This is a communication system that permits interaction with computer using eye movement. The necessary steps of implementation of HCI are EOG signal acquisition and analysis, feature extraction and classification. EOG signal has been acquired by placing electrodes at left and right corner of eye from 12 subjects. Dual Tree Complex Wavelet Transform (DTCWT) has been employed to denoise the EOG signal and 16 features are extracted from the time domain. Three classifiers (Decision Tree (DT), k-Nearest Neighbor (KNN) algorithm, and Support Vector Machines (SVM)) have been used to identify the horizontal eye movement i.e. left and right. Analysis and comparison of their performance is made on the basis of confusion matrix, receiver operating characteristics (ROC) and performance indices i.e. sensitivity, specificity, precision, accuracy and F1 score to evaluate the most efficient classifier for their classification task. According to classification results, out of three classifiers, KNN is the best classifier for horizontal EOG signal and has shown almost 100 percent accuracy.Keywords
This publication has 15 references indexed in Scilit:
- Comparison of decision tree algorithms for EMG signal classification using DWTBiomedical Signal Processing and Control, 2015
- Decision tree structure based classification of EEG signals recorded during two dimensional cursor movement imageryJournal of Neuroscience Methods, 2014
- Automatic EEG seizure detection using dual-tree complex wavelet-Fourier featuresExpert Systems with Applications, 2014
- A polynomial fitting and k-NN based approach for improving classification of motor imagery BCI dataPattern Recognition Letters, 2010
- EOG-based Human–Computer Interface system developmentExpert Systems with Applications, 2010
- The hybrid BCIFrontiers in Neuroscience, 2010
- A comparison of methods for multiclass support vector machinesIEEE Transactions on Neural Networks, 2002
- Complex Wavelets for Shift Invariant Analysis and Filtering of SignalsApplied and Computational Harmonic Analysis, 2001
- Support vector machines for histogram-based image classificationIEEE Transactions on Neural Networks, 1999
- Support-vector networksMachine Learning, 1995