EEG signal classification for BCI applications by wavelets and interval type-2 fuzzy logic systems

摘要: Propose Haar wavelet transformation and ROC curve for EEG signal feature extraction.Combine wavelets interval type-2 fuzzy logic system classification.Benchmark datasets downloaded from the BCI competition II are used experiments.Proposed wavelet-IT2FLS outperforms winner methods of II.IT2FLS dominates competing classifiers: FFNN, SVM, kNN, AdaBoost ANFIS. The nonlinear, noisy outlier characteristics electroencephalography (EEG) signals inspire employment due to its power handle uncertainty. This paper introduces an approach classify motor imagery using (IT2FLS) in a combination with transformation. Wavelet coefficients ranked based on statistics receiver operating characteristic criterion. most informative serve as inputs IT2FLS classification task. Two benchmark datasets, named Ia Ib, brain-computer interface (BCI) II, employed experiments. Classification performance is evaluated accuracy, sensitivity, specificity F-measure. Widely-used classifiers, including feedforward neural network, support vector machine, k-nearest neighbours, adaptive neuro-fuzzy inference system, also implemented comparisons. method considerably comparable classifiers both best Ib reported by 1.40% 2.27% respectively. proposed yields great accuracy requires low computational cost, which can be applied real-time data analysis.