Statistical Feature Analysis of EEG Signals for Calmness Index Establishment
AbstractElectroencephalographic (EEG) signals are very closely related to psychophysiological. The EEG signals displayed few responses which can be categorized. This article discussed the use of statistics over the EEG features which confirm the different mental characteristics. Two different type of stimulus was given named as relaxed state and non-relaxed state. Asymmetry index was computed as the EEG features via the alpha waves and was extracted during the relaxed state and the non-relaxed state. The EEG features were clustered to a group of three, four and five using Fuzzy C-Means. During the relaxed state, the alpha wave showed a higher response as compared to the non-relaxed state. This is observed by using the mean relative energy between the relaxed state and non-relaxed state. To ensure which EEG features in the clusters showed a significant difference, p < .05, a statistical test was used. Wilcoxon Signed Ranks test is the best-statistical test to verify the selected clusters as it is suitable to analyze the small sample of data. Wilcoxon Signed Ranks test used a hypothesis testing which using the same method as paired sample t-test. The advantage in using Wilcoxon Signed Ranks test is that, it uses the median to get the difference between two samples of data. Analytical results showed that the data features of four clusters and three clusters give a significant difference, thus the obtained results can be used to further up the study. The Wilcoxon Signed Ranks test results confirmed that the proposed technique has potential in establishing the calmness index.
Open access licenses
Open Access is by licensing the content with a Creative Commons (CC) license.
This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License.