We found a match
Your institution may have access to this item. Find your institution then sign in to continue.
- Title
Similarity calculation method for time series features based on rough set.
- Authors
Jin Ling; Mu Zhendong
- Abstract
Informatization brings a large amount of time series information, and the method for extracting the feature of such information has been obsolete. In this article, the feature extraction between EEG signals of the same type was carried out by analyzing the EEG signal with an evident time series feature as a study object, by using the rough set thinking to define the upper approximation and lower approximation of a time point feature and via the method for defining similarity. The feature of visual evoked potentials (VEP) of five persons was calculated and verified using definition method. The result shows that the deviations of calculated similarity of 400 samples from each of five subjects are within three; and at the same time the sum of the 400 samples was calculated by taking the superimposed average EEG as a standard. The sample feature of the five subjects was calculated using the method designed herein, and then matching calculation was done with superimposed feature. As a result, the number of samples whose fit is above 0.7 accounts for 80% of the number of all samples.
- Subjects
TIME series analysis; ELECTROENCEPHALOGRAPHY; ROUGH sets
- Publication
Metallurgical & Mining Industry, 2015, Issue 4, p230
- ISSN
2076-0507
- Publication type
Article